Skip to main content

Full text of "General Electric review"

See other formats


- :•:[■-.'■ >i •'- • .;. 
"" ■ ' ■ |J| i " ■' 



Digitized by the Internet Archive 

in 2011 with funding from 

University of Toronto 

General Electric Review 


January, 1915 — December, 1915 





Absolute Zero, The, by Dr. Saul Dushman 93, 238 

Acid-Dipping, Electroplating and Japanning Plant, A 

Modern, by Horace Niles Trumbull 1121 

Agriculture. Electricity in, by C. J. Rohrer 483 

A.I.E.E , History of Schenectady Section of the, by S. M. 

Crego 1006 

A.I.E.E., Notes on the Activities of the. 71, 148, 222, 301, 405. 594 
Air Cleaning Apparatus for the Ventilation of Generators 

and Transformers, by Wm. Baum 801 

Air Supply, Test for Dirt in an, by S. A. Moss 622 

Altitude, Effect of, on the Spark-over Voltages of Bushings, 

Leads and Insulators, by F. W. Peek, Jr 137 

Aluminum Company of America, at Massena Springs, 

N. Y., The 45,000-kw. Synchronous Converter 

Substation of the, by J. L. Burnham and R. C. Muir 873 
Apprentice System at the Lynn Works of the General 

Electric Company, The, by Theodore Bodde 35 

Automobile Industry, Electricity in the, by F. M. Kimball 550 

Ball Bearings in Electric Motors, by F. H. Poor 631 

Berkshire Street Railway, Semi-outdoor Portable Sub- 
station for, by W. D. Bearce 44 

Bethlehem- Chili Iron Mines Company, The 2400- volt 

Railway of the, by E. E. Kimball 12 

Burning Powdered Coal, Some Problems in, by Arthur S. 

Mann 920, 959 

Butte, Anaconda & Pacific Railway, Contact System of the, 

by J B. Cox 842 

Gars, Small, Economies in Operating, by J. F. Layng . . . 790 
Car Operation and Power Consumption, Relation Between, 

by J. F. Layng 973 

Cathode Rays and Their Properties, by J. P. Minton . . . 118 
Cathode Ray Tube and Its Application, The, by M. E. 

Tressler 816 

Cathode Ray Tubes, Some Characteristics of, by J. P. 

Minton 636 

Central Station, The Possibilities Open to the, in Solving 

the Freight Terminal Problem, by Jas. A. Jackson 1142 
Chicago, Milwaukee & St. Paul Railway Company, The 

First 3000-volt Locomotive for the, by E. S. Johnson 1 1.34 
Chicago, Milwaukee & St. Paul Locomotives, The, by 

A. H. Armstrong 600 

Chicago, Milwaukee & St. Paul Railway. The Electrifi- 
cation of the Puget Sound Lines of the, by A. H. 

Armstrong 5 

Chicago, Milwaukee & St. Paul Railway, The 1500-volt 

Electrification of the, by W. D. Bearce 644 

Coal, Powdered, Some Problems in Burning, by Arthur S. 

Mann 920, 959 

Coatings, Protective, for Metals, by H. B. C. Allison 878 

Coffee Plantation, Hydro-electric Installation on a, by 

J. H. Torrens 219 

Cohoes Company at Cohoes, N. Y., Hydro-electric 

Development of the, by B. R. Connell 340 

Columbia University, The New Advanced Course in 

Electrical Engineering at, by W. I. Slichter 940 

Compensators for Mazda C Lamps, by H. D. Brown 596 

Consumer, The Small: A Problem, by A. D. Dudley .... 657 
CONTROL. (See also Protection). 

Sprague-General Electric PC Control, by C. J Axtell 985 
Control and Protection of Electric Systems, by C. P. 

Steinmetz 887 

Coolidge Tube. Application of the. to Metallurgical 

Research, by Dr. Wheeler P. Davey. 134 

Cooling Power Transformers, Principal Factors Governing 

the Choice of Method of, as Related to Their First 

Cost and Operating Conditions, by W. S. Moody. . 839 
Corona and Spark-overin Oil, The Law of, by F. W. Peek, Jr. 821 
Current, Growth of, in Circuits of Negative Temperature 

Coefficient of Resistance, by F. W. Lyle 1129 


Dark Room, X-Ray, A Model, by Wheeler P. Davey . .. 1107 

Depreciation of Property, by W. B. Curtiss 1099 

Developments in Electrical Apparatus During 1914, by 

John Liston 80 

Direct Current, High Potential Methods of Obtaining, 

by Stuart Thomson 1084 

Drill. Rock, The Fort Wayne Electric, by C. Jackson .... 273 

Earth Connections, Proper Construction of, by G. H. 

Rettew 904 

Economies in Operating Small Cars, by J. F. Layng 790 

Electric Power Industry, The, by D. B. Rushmore 427 

Electrical Development, The Trend of, by Paul M. Lincoln 784 
Electrical Engineering at Columbia University, The New 

Advanced Course in, by W. I. Slichter 940 

Electricity in the Automobile Industry, by Fred M. 

Kimball 550 

Electricity in the Construction and Operation of the 

Panama Canal (Supplement to July issue), by 

Edward Schildhauer 679 

Electriquette. The Osborne, by O. E. Thomas 299 

Electricity in Agriculture, by C. J. Rohrer 483 

Electro-culture: A Resumed of the Literature, by Helen R. 

Hosmer 14 

Electron Discharge, The Pure, and Its Applications in 

Radio Telegraphy and Telephony, by Irving 

Langmuir 327 

Electroplating, Acid-Dipping and Japanning Plant, A 

Modern, by Horace Niles Trumbull 1121 


Application of the Electron Theory to Various 

Phenomena, by J. P. Minton 287 

Electromagnetic Radiation from the Viewpoint of the 

Electron Theory, by J. P. Minton 387 

Cathode Rays and Their Properties, by J. P. Minton 118 
Electron Theory of Electric Conduction in Metals, by 

J. P. Minton 204 

Some Characteristics of Cathode Ray Tubes, by J. P. 

Minton 636 

Emergency Transformer Connections, by George P. Roux 832 
Enameling Ovens, Electrically Heated, by C. W. Bartlett L130 
Engines, Internal Combustion, Parallel Operation of 

Alternating Current Generators Driven by, by 

R. E. Doherty and H. C. Lehn 167 

Engineer, The Status of the, by Dr. E. W. Rice, Jr 234 

Engineering in the Navy, by W. L. R. Emmet 1097 

Eye and Illumination, The, by H. E. Mahan 268 

Factory Lighting, by G. H. Stickney 67 

Factories, Isolated Power-house for, by W. E. Francis. . . 1057 
Fire Departments, Portable Searchlights for, by L. C. 

Porter and P. S. Bailey 1144 

Freight Terminal Problem. The Possibilities Open to 

Central Stations in Solving the. by Jas. A. Jackson 1142 
Frequency Changers, Parallel Operation of, by G. H. Rettew 836 

Gases, The Kinetic Theory of, by Dr. Saul Dushman . . . 

952, 1042, 1159 

Gases, Noble, Notes on the, by W. S. Andrews 226, 40S 

Gears and Pinions, Railway Motor, Operating Conditions 

of, by A. A. Ross 249 

Genemotor, The, by M. J. Fitch 384 

General Electric Company's Exhibits at the Panama- 
Pacific International Exposition, by G. W. Hall. . . 561 


Air Cleaning Apparatus for the Ventilation of Gener- 
ators and Transformers, by Wm. Baum 801 

Parallel Operation of Alternating Current Generators 
Driven by Internal Combustion Engines, by R. E. 

Doherty and H. C. Lehn 167 

Grounding, General Notes on, by H. M. Wolf 991 


Heating and Heating Appliances, Electric, by C. P. 

Randolph s23 

High Frequency, by F. W. Peek, Jr 934 

High Potential Direct Current. Methods of Obtaining, by 

Stuart Thomson 1084 

High Voltage Direct-current Substation Machinery, by 

E. S. Johnson 641 

High Voltage Arrester for Telephone Lines, by E. P. Peck 189 

Hoists, Large Steam. Tests of, by H. E. Spring 179 

"Home Electrical" at the Panama-Pacific International 

E xposition. The. by Don Cameron Shaffer 572 


Hydro-electric Development of the Cohoes Company. 

at Cohoes. N. Y„ by B. R. Connell 340 

Water Powers of New England, by H. I. Harriman . . 358 
Hydro-electric Installation on a Coffee Plantation, by 

J. H. Torrens 219 

Impregnating Cods, The Process of, and a Large Impreg- 
nating Plant, by Robert Reid 48 

Industrial Applications of Electricity, Some, by A. R. 

Bush 46D 

Industrial Research, by L. A. Hawkins 416 

Industry, The Electric Power, by D. B. Rushmore .... 4L'7 
Industry. The Individual and Corporate Development of. 

by C. P. Steinmetz 813 

Insulating Materials, The Volume Resistivity and Surface 

Resistivity of, by Harvey L. Curtis 996 

Insulations, Solid. Electrical Characteristics of, by F. W. 

Peek, Jr 1050 

Insulation Testing, by G. B. Shanklin 1008 

Iron-cobalt Alloy. FeaCo, and Its Magnetic Properties, The. 

by Trygve D. Yensen 881 

Japanning. Acid-Dipping and Electroplating Plant, A 

Modern, by Horace Niles Trumbull 1121 

Jitney Problem, The, by J. C. Thirlwall 604 

Kenotron. The. A New Device for Rectifying Alternating 

Currents, by Dr. Saul Dushman 156 

Kinetic Theory of Gases, The, by Dr. Saul Dushman 

952, 1042, 1159 

Compensators for Mazda C Lamps, by H. D. Brown 596 

Electric Lamp Industry, by G. F. Morrison 497 

High Candle-power Mazda Lamps for Steel Mill 

Lighting, by G. H. Stickney 377 

Incandescent Lamps for Projectors, by L. C. Porter. . 371 
Modern Street Lighting with Mazda Lamps, by H. A. 

Tinson 659 


Brief Review of the Electric Lighting Industry, by 

C. W. Stone 439 

Eye and Illumination. The, H. E. Mahan 268 

Factory Lighting, by G. H. Stickney 67 

High Candle-power Mazda Lamps for Steel Mill 

Lighting, by G. H. Stickney 377 

Illumination of the Panama-Pacific International 

Exposition, by W. D'A. Ryan 

Lighting of Ships. The, by L. C. Porter 143 

rn Street Lighting with Mazda Lamps, by H. A. 

Tinson 659 

Sign and Building Exterior Illumination by Projection, 

by K. W. Mackall and L. C. Porter 2S2 

Lock Entrance Caisson for the Panama Canal, by L. A. 

Mason 210 


Chicago, Milwaukee & St. Paul Locomotives, The, 

by A. H. Armstrong 600 

Firs' 300(1- volt Locomotive for the Chicago, Milwaukee 

& St. Paul Railway Company. The. by E. S. Johnson 1154 
Operation and Rating of the Electric Locomotive. The. 

by A. H. Armstrong 828 

Towing Locomotives for the Panama Canal. The. 
by C. \V. Larson iQl 

Lubrication. The Theory of, by L. Ubbelohde. Translated 

by Helen R. Hosmer 966. 1074. 111S 

Magnetic Properties of Steel, The Effect of Chemical 

Composition Upon the, by W. E. Ruder 19? 

Magnetization Curves, Some Notes on, by John D. Ball 31 

Marine Work. Electricity in, by Maxwell W. Day 504 

Metal, Protective Coatings for, by H. B. C. Allison 878 

Metals, Radiography of, by Wheeler P. Davey 795 

Meter Design, Induction, Some Notes on, by W. H. Pratt 277 

Mica, Built Up, X-Ray Examination of, by C. N. Moore 195 

Mine Haulage Motor, The Modern, by C. W. Larson. . . . 264 

Mining Work. The Use of Electricity in, by D. B. Rushmore 527 
Motion Picture Machines, Current Supply for. by H. R. 

Johnson 895 


Electric Motor in the Printing Industry. The, by 

W. C. Yates 1136 

Ball Bearings in Electric Motors, by Frederick H. Poor 631 
Methods of Removing the Armature from Box Frame 

Railway Motors, by J. L. Booth 908 

Modern Mine Haulage Motor, The, by C. W. Larson 264 
Power Consumption of Railway Motors, by H. L. 

Andrews and J. C. Thirlwall 944 

Railway Motor Characteristic Curves, by E. E. 

Kimball 296 

Short Method for Calculating the Starting Resistance 
for Shunt, Induction and Series Motors, A. by 

B. W. Jones 131 

Subdivision of Power as Solved by the Small Motor, 

The, by R. E. Barker and H R. Johnson 555 


Electrical Equipment of the Vermont Marble Com- 
pany, by John Liston 1015 

Electricity in Agriculture, by C. J. Rohrer 483 

Electricity in Marine Work, by Maxwell W. Day .... 504 

Electric Power in the Textile Industry, by C. A. Chase 540 
Industrial Applications of Electricity, Some, by A. R. 

Bush 460 

Supplying of Power to the Quaker Oats Company, by 

J. M. Drabelle 42 

Use of Electricity in Mining Work. The. by D. B. 

Rushmore 527 

Multi-recorder, A Cursory Account of the First Lightning 
Storm of the Season as Given by the Records of the, 

by E. E. F. Creighton 860 

Navy, Engineering in the, by W. L. R. Emmet 1097 

Negative Temperature Coefficient of Resistance. Growth 

of Current in Circuits of, by F. W. Lyle 1129 

N.E.L.A. Lamp Committee Report, A Review of the, by 

G. F. Morrison 925 

New England, Water Powers of, by H. I. Harriman 358 

Noble Gases, Notes on the, by W. S. Andrews 226, 408 


In Memoriam: Douglas S. Martin 76 

In Memoriam: John P. Judge 672 

In Memoriam: Dr. and Mrs. F. S. Pearson 930 

In Memoriam: George Crellin Cartwright 1169 

Oil, The Law of Corona and Spark-over in. by F. W. Peek, Jr. 821 

Ontario Municipal Railway, The 1500-volt Direct-current 

Electrification of the, by G. H. Hill 10 

Oscillations, Damped, the Production of, by Leslie O. 

Heath 1110 

Ovens, Enameling. Electrically Heated, by C. W. Bartlett 1130 

Panama Canal. Electricity in the Construction and Opera- 
tion of the. by Edward Schildhauer 679 

Ancon Quarry 688 

Balboa Sand Dock 688 

Control of the Lock Machinery 748 

Distribution at Locks ■ 716 

Gatun Hydro-electric Station 688 

Gatun Locks and Dam 679 

Interlocking (Panama Canal Lock) 754 


Panama Canal, Locomotive Design, Details of 732 

Machinery for the Operation of the Locks and Spillways 722 

Pacific Locks and Dam 687 

Reserve Station 70S 

Towing Locomotives 729 

Transmission System 709 

Panama Canal, Lock Entrance Caisson for the, by L. A. 

Mason 210 

Panama Canal, The Towing Locomotives for the, by C. W. 

Larson 101 

Panama-Pacific International Exposition, The Illumination 

of the, by W. D'A. Ryan 579 

Panama-Pacific International Exposition, The General 

Electric Company's Exhibits at the, by G. W. Hall 561 

Panama- Pacific International Exposition, The "Home 

Electrical " at the, by Don Cameron Shafer 572 

Parallel Operation of Alternating-current Generators 
Driven by Internal Combustion Engines, by R. E. 

Doherty and H. C. Lehn ' 167 

Parallel Operation of Frequency Changers, by G. H. Rettew 836 

Paths of Progress. The, (Editorial) 

3, 79, 155, 231, 314, 415, 599. 783. 867, 939, 1011, 1091 

Periodic Law, The, by Dr. Saul Dushman 614 

Pinions (See also Gears) 

Power, Subdivision of, as Solved by the Small Motor, by 

R. E. Barker and H. R. Johnson 555 

Power Consumption of Railway Motors, by H. L. Andrews 

and J. C. Thirlwall 944 

Power House, Isolated, for Factories, by W. E. Francis. . 1057 

Practical Experience in the Operation of Electrical 
Machinery, by E. C. Parham 

Alternator Speed Low 58 

Armature Threw Solder 1082 

Belts, Loose 58 

Brush-holders Shifted 59 

Burn-out due to Core Loss 57 

Brake Adjustments, Electric 217 

Capacity Current 40 1 

Clutches. Adjusting Single-phase Motor 929 

Commutators, Loose 56 

Commutator Winding, Improvised 403 

Connection, Loose 5G 

Contact-shoe Pressure, Excessive 217 

Crane Troubles. . 1003 

Core, Loose 5S 

Deflections, Misleading . 402 

Devices, Misapplication of 401 

Elevator Trouble 1 155 

Elevator Speed, Erratic 59 

Equalizer on the Wrong Side 305 

Field Connection Error 928 

Generators Motoring at No-Load . . 305 

Hot Box Indications 57 

Instrument Connections Wrong 404 

Load was Unbalanced. 1082 

Motor Acceleration, Jerky 218 

Motor Heating, Repulsion 667 

Motor Mounting, Changing 306 

Motors, Repulsion Induction, Heating and Sparking of, 147 

Motor Reversed 115o 

Motor Stopped and Reversed 1004 

Motor Throwing Oil 307 

Motor on an Inertia L^ad, Variable-Speed 667 

Motor Would Not Start 928 

Power-Factor, Low 56 

Pump Output, Excessive 147 

Reactor Starting-Box Trouble 403 

Repulsion-induction Motors, HeatMig and Sparking of 147 

Resistance Wire Crossed 1154 

Rotor Rubbed Stator 218 

Shunt Ratio, The Wrong 666 

Slip-ring Contacts. Imperfect - . .. 304 

Stations in Series 861 

Stator Coil Connections 1083 

Transformer Connections SOU 


Practical Experience, etc., by E. C. Parham -Cont'd 

Transformer Failures j^g 

Transformers, Parallel ggj 

Voltage, Service, Too Low gg x 

Voltage, Unstable | q, , -, 

Printing Industry, The Electric Motor in the, by W. C. 

Yates 1136 

Projectors, Incandescent Lamps for, by L. C. Porter 371 

Protection and Control of Industrial Electric Power, by 

C. P. Steinmetz gyg 

Protection of Railway Signal Circuits against Lightning 

Disturbances, by E. K. Shelton 1127 

Quaker Oats Company, The Supplying of Power to the, 

by J. M. Drabelle 42 

Radiography of Metals, by Dr. Wheeler P. Davey 795 

Radio-telephony, by W. C. White 33 

Radio Telegraphy and Telephony, The Pure Electron 
Discharge and Its Application In, by Irving 

Langmuir 327 

Railways, Electric, A Review of, by W. B. Potter and 

G. H. Hill 444 


Automatic Railway Substations, by Cassius M. Davis 976 
Contact System of the Butte, Anaconda & Pacific 

Railway, by J. B. Cox 842 

Operating Conditions of Railway^Motor Gears and 

Pinions, by A. A. Ross 249 

Selection of Railway Equipment, The, by J. F. Layng 126 
Railway Motor Characteristic Curves, by E. E. 

Kimball 296 

Sprague-General Electric PC Control, by C. J. Axtell 985 
Rectifying High Tension Alternating Currents, A New 

Device for, by Dr. Saul Dushman 156 

Refrigeration Field as It Exists Today, A Survey of the, by 

H. I. Hollman 65 

Refrigeration, A Standard in, by L. A. Simmons 1170 

Research, by Dr. W. R. Whitney 1012 

Research, The Relation of, to the Progress of Manu- 
facturing Industries, by Dr. W. R. Whitney 868 

Research, Industrial, by L. A. Hawkins 416 

Resistance Standards, Precision. Ten-to-one Ratio for 

Comparing, by C. A. Hoxie 915 

Resistance, Starting, for Shunt, Induction, and Series 
Motors, A Short Method for Calculating the, by 

B. W. Jones 131 

Resistivity, Volume, and Surface Resistivity of Insulating 

Materials, The, by Harvey L. Curtis 996 

Resolutions Presented to C. A. Coffin and E. W. Rice, Jr.. 
by Association of Edison Illuminating Companies, 
Reproduction of the (Supplement to March issue). 

Rheostats. Water, by X. L. Rca 1001 

Rock Drill, The Fort Wayne Electric, by C. Jackson 273 

Searchlights, Portable, for Fire Departments, by L. C. 

Porter and P. S. Bailey 1 144 

Ships, The Lighting of, by L. C. Porter 143 

Short Circuits, Electrical, Mechanical Effects of, by S. H. 

Weaver 1066 

Sign and Building Exterior Illumination by Projection, by 

K. W. Mackall and L. C. Porter 282 

Signal Circuits, Railway, Protection of. Against Lightning 

Disturbances, by E. K. Shelton 112/ 

Slot Insulation Design, Some Aspects of, by H. M. Hobart 360 

Status of the Engineer, The, by E. W. Rice, Jr 234 

Steel Castings, An X-Ray Inspection of, by Dr. Wheeler P. 

Davey 25 

Steel, Magnetic Properties of. The Effect of Chemical 

Composition Upon the, by W. E. Ruder 197 

Steel Mill Lighting, High Candle-power Mazda Lamps for, 

by G. H. Stickney 377 

Spark-over Voltages of Bushings, Leads and Insulators, 

Effect of Altitude on the, by F. W. Peek, Jr 137 

Sprague-General Electric PC Control, by C. J. Axtell. . . . 985 


Automatic Railway Substations, by Cassius M. Davis 976 
45.000-kw. Synchronous Converter Substation of the 
:r.inum Company of America at Massena Springs. 
N. V.. The. by J. L. Burnham and R. C. Muir. . . . S73 
High Voltage Direct-current Substation Machinery, 

by E. S. Johnson 641 

Semi-outdoor Portable Substation for Berkshire Street 

Railway, by W. D. Bearce « 

"Supplies": Devices and Appliances for the Distribution. 

Control and Utilization of Electricity, by S. H. Blake 553 
Switchboard Apparatus. Some Recent Developments In. 

by E. H Beckert 646 

Telephone Lines. High Voltage Arrester for, by E. P. Peck 189 
Temperature Coefficient Formula; for Copper, by John D. 

Ball 669 

- Dirt in an Air Supply, by S. A. Moss 622 

The High Tension . by Wm. P. Woodward 39S 

Tests. Electrical. Made in 1883 and Their Influence on 

Modern Testing. A Series of. by A. L. Rohrer 22 

Tests of Large Steam Hoists, by H. E. Spring 179 

Textile Industry. Electric Power in the, by C. A. Chase . .540 
Thury System of Direct-current Transmission. The, by 

Wm. Baum 1026 


nomies in Operating Small Cars, by J F. Layng. 790 
Electrification of the Puget Sound Lines of the 
Chicago. Milwaukee & St. Paul Railway, The. by 

A. H. Am-strong 5 

Relation Between Car Operation and Power Con- 
sumption, by J. F. Layng 973 

Review of Electric Railways. A. by W. B. Potter and 

G. H. Hill 444 

Semi-outdoor Portable Substation for Berkshire Street 

Railway, by W. D. Bearce 44 

2400-volt Railway of the Bethlehem-Chili Iron Mines 

Company. The. by E. E. Kimball 12 

1 500-voIt Direct-current Electrification of the Ontario 

Municipal Railway. The. by G. H. Hill 10 

1500-volt Electrification of the Chicago. Milwaukee 
& St. Paul Railway, by W. D. Bearce 644 



Air Cleaning Apparatus for the Ventilation of Gener- 
ators and Transformers, by Wm. Baum 801 

Emergency Transformer Connections, by George P. 

Rous 832 

High Potential Transformer Testing Equipment, by 

Wm. P. Woodward 398 

Mechanical Stresses in Shell Type Transformers, by 

J. Murray Weed 60 

- on the Operation of Transformer used with 

2-ka-., 100.000-cycle Alternator, by S. P. Nixdorff 308 
Open-delta or V Connection of Transformers, by 

George P. Rous 52 

Principal Factors Governing the Choice of Method of 

Cooling Power Transformers as Related to Their 

First Cost and Operating Conditions, by W. S. 

Moody 839 

Transients, The Infinite Duration of, by Charles L. Clarke 73 
Transmission Line Calculator, A, by Robert W. Adams. . 28 

Electric Transmission of Power, by R. E. Argersinger 454 
Theory of Electric Waves in Transmission Lines, by 

J. M. Weed 1148 

Thury System of Direct Current Transmission, The. 

by Wm. Baum. . , 1026 

Wireless Transmission of Energy, by Elihu Thomson 316 

Ventilation of Generators and Transformers, Air Cleaning 

Apparatus for the, by Wm. Baum 801 

Vermont Marble Company, Electrical Equipment of the, 

by John Liston 1015 

Waves. Electric. Theory of in Transmission Lints, by 

J. M. Weed 1148 

Welfare Work, by Jesse W. Lilienthal 1092 

Wireless Transmission of Energy, by Elihu Thomson 316 

X-rays, by Dr. Wheeler P. Davey 258, 353, 625 

X-ray Dark Room, A Model, by Wheeler P. Davey .... 1107 
X-ray Examination of Built-up Mica, by C. N. Moore. . . 195 
X-ray Inspection of a Steel Casting, An, by Dr. Wheeler P. 

Davey 25 

X-rays. Some Notes on, by W. S. Andrews 152 

Zero. The Absolute, by Dr. Saul Dushman 93. 238 


1913 — 1914 — 1915 


Charging. . No. 79 

Charging resistance; Function of ... . No. 39 

Desirability for steel mill circuits . . . Xo. S8 

loisture in Xo. 44 


Battery auxiliary vs. overload capac- 
ity of d-c. generator Xo. 82 

Peak load in a-c. installations; Suit- 
ability to carry 

Position on side of pole; Choice of, . No. 115 

Location on d-c. dynamos Xo. 60 


Mounting, type depended 

and losses 
various arrangem <.-: 
ng capacity 



(1914) 159 


(1914) 338 

(1914) 159 

(1914) 1002 

(1913) 1001 


(1915) 410 

CABLE— Cont'd 

Pot heads; Necessity for 

Size for a certain installation 

Varnished cambric, advisability of 

using this type in ducts 


Concentrated beam of light; Suit- 
ability to use as a measure of 



Current conduction when copper 


Polarization reduced by zinc amalgam 


Iron vs. air core 

Reactance formula for various shapes 

Polarity, testing while on machine. . . 

Short circuits in field winding of 

railway motor; Detection of 

Q. &. A. 



Xo. 156 



Xo. 81 



Xo. 74 



Xo. 36 



No. 14 

Xo. 49 

Xo. 84 
No. 102 

No. 42 
Xo. 150 

(1913) 275 

(1913) 539 

(1914) 160 

(1914) 507 

(1913) 537 

(1915) 1086 

COILS — Cont'd Q- & A. YEAR PAGE 


Current division between two parallel 
circuits interconnected by coil; 

Calculation of No. 58 (1913) 683 

110.000-volt coil; Impracticability of 

constructing a No. 154 (1915) 1168 

Protection by a coil automatically 

inserted in a line; Lack of No. 154(1915) 1168 

Reactance resulting from varied com- 
bination of coils; Method of cal- 
culating No. 11 (1913) 274 

(And relays). Troubles on lines; 

Practicability of segregating No. 146 (1915) 935 


Grooving; Reasons for No: 2 (1913) 207 


Area and pressure, relative electrical 

importance of each No. 43(1913) 538 


Brushes raised at starting a commu- 

tating pole machine, reasons No. 77 (1914) 80 

Connections and unbalanced three- 
wire d-c. load No. 142 (1915) 670 

Line drop; Limit and effect of No. 109 (1914) 772 

Neutral for three-wire d-c. line de- 
rived from machine's step-down 

transformers No. 89(1914) 338 

Polyphase machine operating single- 
phase; Effect of No. 118 (1914) 1003 

Shunt around commutating pole 
winding should be inductive; 

Reasons why No. 29 (1913) 463 


Insulation; Effect on No. 63 (1913) 999 


Charged dust particles; Effect of 

direct current on No. 103 (1914) 508 

Earth; Possibility of obtaining from. No. 100 (1914) 506 

Substitute for standard material ... . No. 25 (1913) 344 

Hunting caused by relation of gover- 
nor to automatic voltage regulator 

on driven generator No. 55 (1913) 612 


Control by automatic voltage regu- 
lator No. 98 (1914) 606 

Driving methods No. 16(1913) 276 


Trolley circuit considerations No. 119 (1914) 1003 


Checking in electrical machines; 

Methods of No. 38 (1913) 466 


Advantages and disadvantages of 
synchronous motor and induction 
motor driven sets for tying-in two 

systems No. 51(1913) 609 


Construction, special design No. 71 (1913) 1002 

Melting of non-ferrous metals; Ref- 
erences on No. 66 (1913) 1000 


Forced-draft ventilation for low-speed 

machines No. 157 (1915) 1 169 


Armature reconnection for a different 

voltage; Possibility of a certain. .. . No. 92 (1914) 428 

Bearing current; Explanation of No. 26 (1913) 344 

Bearing current; Detection and meas- 
urement of No. 136 (1915) 311 

GENERATOR — Cont'd Q- & A. year page 


Control of two paralleled machines 

individually by twe automatic volt- „ T 

age regulators j No ' 107 < 1914 > > 71 

iNo. 116 (1914) 1002 

Coupling two machines mechanically 

to run in parallel No. 91 (1914) .(40 

Flux; Full-load value relative to no- 
load value of No. 106 (1914) 771 

Induction machines driven by low- 
pressure steam turbines No. 96 (1914) 431 

Overheated solid core when three- 
phase machine runs single-phase. . . No. 132 (1915) 228 

Power-factor on short circuit No. 54 (1913) 612 

Power-factors 70 and 100 per cent, 

difference in input No. 143 (1915) 671 

Regulation; Question of improving. . No. 20 (1913) 343 

Wave shape of inductor type machine No. 56 (1913) 683 

High-voltage machines existent and 

design limitations No. 140 (1915) 410 

Load divided disproportionately 

between two paralleled machines. . No. 3 (1913) 207 

Overload capacity vs. storage battery 

auxiliary No. 82 (1914) 159 

Shunt around commutating pole 
winding should be inductive; Rea- 
sons why No. 29 (1913) 163 

Voltage of two machines in series, 
difficulty in maintaining on increase 
of load No. 47 (1913) 638 

Voltage regulation of automobile 
lighting generators by third-biush 

method No. 114 (1914) 932 


End-thrust, possible effects when un- 
balanced No. 105 (1914) 508 

Integral vs. external fan ventilation . . No. 110 (1914) 772 

National Electrical Rules for neutral . No. 125 (1915) 74 

Street lighting circuits, protection 

against grounding by trees. ....... No. 33 (1913) 464 


Definition and testing for presence. . . No. 147 (1915) 936 

Breakdown voltages compared with 

those of needle-gap No. 52(1913) 611 


Corona 's effect No. 63 (1913) 999 


Leakage current, its nature, and why 

it takes place No. 59 (1913) 755 


Inter-relationship No. 151 (1915) 1086 


Efficiency of carbon and tungsten 

incandescent types No. 83 (1914) 160 


Mines supplied from 230-v. taps of 

4600/2300-v. transformer No. 27 (1913) 463 

Voltage regulation of automobile gen- 
erators by third-brush method .... No. 114 (1914) 932 

Current division between two parallel 
circuits interconnected by react- 
ance coil No. 58 (1913) 683 

Multiple vs. single No. 65 (1913) 1000 

Reactance of three wires in a plane. . No. 28 (1913) 463 

Sag and size of conductor No. 139 (1915) 410 


Steel mill; Average running load for. . No. 80 (1914) 159 
Three-wire circuit ;Calculation for a. . No. 69 (1913) 1001 



Construction and uses for which it is 

especially suited No. 10 (1913) 274 


Curve-drawing, connections No. 75 (1914) 80 

Reversal of one on low power-factor 
when two are measuring three- 
phase power No. 4 (1913) 208 

Power-factor of three-phase line ob- 
tained from ratio of two readings. . No. 35 (1913) 465 


Frequency, effect of change on ac- 
curacy No. 99 (1914) 506 

Protection against lightning No. 73 (1913) 1002 


Three-phase power; Explanation of 

two-meter method of measuring. . . No. 48 (1913) 539 


Square and circular mils; Difference 

between and method of calculating No 7 (1913) 208 


Explosion-proof types; Construction 

of No. 21 (1913) 343 

Open and enclosed types; Definition 
of. (Later, See: Standardization 
Rules of the A.I.E.E. edition of 
Feb. 1. 1915, 5§ 160-172) No. 86 (1914) 338 

Output of d-c. machine, proof that 
maximum occurs when loaded to 

half speed No. 34 (1913) 465 


Brass rs. fiber slot wedges for holding 

in coils No. 41 (1913) 537 

Dynamic braking of squirrel-cage 
type by application of direct cur- 
rent to stator No. 68 (1913) 1001 

Generator; Ability to act as a No. 104 (1914) 508 

Half-voltage; Characteristics at No. 93 (1914) 428 

Knocking sound No. 97 (1914) 431 

Low-speed type; Characteristics of. . No. 108 (1914) 771 

Low-voltage; Characteristics as af- 
fected by No. 50 (1913) 539 

Phase-wound rotor type; Relation of 

heating to speed of No. 57 (1913) 683 

Poles, change in number limited by 

certain factors No. 145 (1915) 864 

Quarter-phase to three-phase recon- 

nection No. 153 (1915) 1168 

Rotor-bar insulation charred, its 

effect on machine's characteristics. No. 126 (1915) 74 

Rotor-bar insulation charred, its re- 
pair and effect on machine's char- 
acteristics . No. 137 (1915) 409 

Starting difficulty No. 72 (1913) 1002 

Three-phase machine operating on 

two-phase circuit No. 1 (1913) 207 

Twenty-five cycle machine operating 

on 60-cycle supply . No. 101 (1914) 506 

Unbalanced phase voltages; Heating 

of three-phase machine on No. 135 (1915) 311 

Unbalanced phase voltages; Heating 

of two-phase machine on No. 128 (1915) 75 


Commutator bars burned as a result 

of reversed armature coil No. 12 (1913) 275 

Field coils; Detection of short cir- 
cuits in . . . No. 150 (1915) 1086 

Low voltage a cause of increased de- 
terioration in mining locomotives. . No. 120 (1914) 1003 

Power-factor; Calculation of improve- 
ment produced in No. 37 (1913) 466 

MOTOR— Cont'd Q- * A * 


Power-factor ; Explanation of influ- 
ence on No. 46 


Delta connected transformers; Method 

of bringing out from No. 149 

National Electrical Rules on ground- 
ing of No. 125 


Concentration, Degree of No. 62 

Respiration; Effect on No. 6 


Combinations that are possible in 

connecting two three-phase lines. . No. 78 
Voltage measurements between two 

lines, peculiar readings Xo. 121 


Definitions and determination No. 144 


Reduction by zinc amalgam in pri- 
mary cells No. 49 


Wet and dry process product; Char- 
acteristics of No. 17 


Combination of several ; Calculation ot No. 8 
Improvement by synchronous motor; 

Calculation of No. 37 

Synchronous motor influence No. 46 

Three-phase value obtained from 

ratio of two wattmeter readings. . . No. 35 

Automatic Voltage 

A-c. to d-c. operation; Change from. . No. 117 
Control of two paralleled a-c. genera- „ _._,_ 

tors individually by two regulators \* ' t ,_ 

LNo. lib 

Exciter controlled by No. 98 

Hunting caused by relation to gover- 
nor on engine driving generator. . . No. 55 

Three-phase unit operating single- 
phase No. 87 


Rupturing capacity of oil switch; Ef- 
fect of time limit on No. 30 


Temperature coefficient of copper. . . No. 113 
Transformer windings; Measurement 

of No. 129 


Function of as applied to lightning 

arrester No. 39 

Field Discharge 

Action; Explanation of No. 23 


Effectiveness of protection No. 90 


Control of electric lamps No. 13 


Bafflers; Purpose of No. 95 

Connections of circuit-opening equip- 
ment No. 94 

Direct current; Utilization on No. 32 

Interrupting action; Explanation of. No. 9 
Rupturing capacity and time-limit 

relay No. 30 


Kelvin scale, basis and layout No. 141 





























































Boosting with an ordinary single- 
phase unit No. 

Breakdown due to high electrostatic 
stress on a certain grounded neutral 
circuit No. 

Burnout; A peculiar No. 

Division of load between two paral- 
leled units No. 

Exchange current between two paral- 
leled units No. 

Internal explosion, cause and preven- 
tion No 

Lighting of mine by 230-v. tap on 
4600/2300-v. unit No. 

Neutral brought out from delta con- 
nected secondaries No. 

Overheating of one delta leg No. 

Overheating of one paralleled with 
another No. 

Parallel operation of two banks, con- 
nections Y-delta, delta-delta No. 

Phasing-out of small polyphase units. No. 

Phasing-out of large three-phase units No. 

Power-factor when short circuited . . . No. 

Ratio change by bringing out a tap. . No. 

Regulation; Method and example of ... 

114." i N0 " 

calculating s , T 

I No. 



155 (1915) 


40 (1913) 
24 (1913) 


22 (1913) 


31 (1913) 


122 (1914) 


27 (1913) 


149 (1915) 
131 (1915) 


45 (1913) 


61 (1913) 

112 (1914) 

111 (1914) 

76 (1914) 

70 (1913) 






5 (1913) 
53 (1913) 




Resistance measurement of windings. No. 129 (1915) 75 
Two-phase to three-phase transfor- 
mation, per cent taps and vectors . . No. 130 (1915) 227 
Two-phase to three-phase transfor- 
mation with three units No. 133 (1915) 310 

Twenty-five cycle unit operating on 

60-cycle supply No. 64 (1913) 1000 


Advantages to be gained from their 

employment No. 18 (1913) 276 

Leads (current and potential) in the 

same conduit No. 127 (1915) 75 

Protection against lightning .. No. 73 (1913) 1002 


Open circuited secondary, reason for 

excessive voltage rise No 15 (1913) 276 


Relief valve on low-pressure end .... No. 124 (1914) 1230 

Alternating current inapplicable No. 134 (1915) 310 


Advantages and properties of this 

type of insulation No. 19 (1913) 343 


Catenary suspension; Stress formulae 

for No. 123 (1914) 1230 



Adams, Robert W. 

Transmission Line Calculator 28 

Allison, H. B. C. 

Protective Coatings for Metal 878 

Andrews, W. S. 

Notes on the Noble Gases 226, 408 

Some Notes on X-rays 152 

Andrews, H. L. 

Power Consumption of Railway Motors 944 

Argersinger, R. E. 

Electric Transmission of Power 454 

Armstrong, A. H. 

Chicago, Milwaukee & St. Paul Locomotive, The. . . . 600 

Electrification of the Puget Sound Lines of the Chicago, 

Milwaukee & St. Paul Railway, The 5 

Operation and Rating of the Electric Locomotive, 

The 828 

Axtell, C. J. 

Sprague-General Electric PC Control 985 

Bailey, P. S. 

Portable Searchlights for Fire Departments 1144 

Ball, John D. 

Temperature Coefficient Formula? for Copper 669 

Some Notes on Magnetization Curves 31 

Baker. R. E. 

Subdivision of Power as Solved by the Small Motor, The 555 
Bartlett, C. W. 

Electrically Heated Enameling Ovens 1130 

Baum, Wm. 

Air Cleaning Apparatus for the Ventilation of Gener- 
ators and Transformers 801 

Thury System of Direct-current Transmission, The. . 1026 
Bearce, W. D. 

Semi-outdoor Portable Substation for Berkshire Street 

Railway 44 

1500-volt Electrification of the Chicago, Milwaukee 

& St. Paul Railway, The 644 

Beckert, E. H. 

Some Recent Developments in Switchboard Apparatus 646 



Blake, S. H. 

"Supplies:" Devices and Appliances for the Dis- 
tribution, Control and Utilization of Electricity. . . 553 
Bodde, Theodore 

Apprentice System at the Lynn Works of the General 

Electric Company, The 35 

Booth, J. L. 

Methods of Removing the Armature from Box Frame 

Railway Motors 90S 

Brown, H. D. 

Compensators for Mazda C Lamps 596 

Burnham, J. L. 

45,000-Kw. Synchronous Converter Substation of the 
Aluminum Company of America at Massena 

Springs, The 873 

Bush, A. R. 

Some Industrial Applications of Electricity 460 

Chase, C. A. 

Electric Power in the Textile Industry 540 

Clarke, Charles L. 

Infinite Duration of Transients, The 73 

Connell, B. R. 

Hydro-electric Development of the Cohoes Company 

at Cohoes, N. Y., The 340 

Cox, J. B. 

Contact System of the Butte, Anaconda & Pacific 

Railway, The 842 

Crego, S. M. 

History of Schenectady Section of the A.I.E.E 1006 

Creighton, E. E. F. 

Cursory Account of the First Lightning Storm of the 
Season as given by the Records of the Multi-recorder 860 
Curtis, Harvey L. 

Volume Resistivity and Surface Resistivity of In- 
sulating Materials, The 996 

Curtiss, W. B. 

Depreciation of Property 1099 

Davis, Cassius M. 

Automatic Railway Substations 760 

Davey. Wheeler. P., Dr. 

Application of the Coolidge Tube to Metallurgical 

Research 134 

Model X-Ray Dark Rwm. A 1107 

Radiography of Metals 7fl5 

X. r ays 258. 353, 625 

X-ray Inspection of a Steel Casting. An . 25 

Day. Maxwell W. 

Electricity in Marine Work 504 

Doherty. R. E. 

Parallel Operation of Alternating Current Generators 

Driven by Internal Combustion Engines 167 

DrabeUe. J. M. 

Supplying of Power to the Quaker Oats Company. The 42 
Dushman, Saul, Dr. 

Absolute Zero. Th( 93. 238 

Kinetic Theory of Gases. The 952, 1042. 1159 

New Device for Rectifying High Tension Alternating 

Currents, A 156 

Periodic Law. The. . . 614 

Dudley. A. D. 

Small Consumer, The: A Problem 657 

Emmet. W. L. R. 

Engineering in the Navy - . ■ 1097 

Fitch. M. J. 

Genemotor, The 384 

Francis, W. E. 

Isolated Power House for Factories 1057 

Hall. G. W. 

General Electric Company's Exhibits at the Panama- 
Pacific International Exposition. The 561 

Harriman, H. I. 

Water Powers of Xew England 358 

Hawkins. L. A. 

Industrial Research 416 

Heath. Leslie O. 

Production of Damped Oscillations. The 1110 

Hill, G. H. 

1500-volt Direct-current Electrification of the Ontario 

Municipal Railway. The 20 

Review of Electric Railways. A. . . 444 

Hobart. H. M. 

Some Aspects of Slot Insulation Design 366 

Hollman. H. I. 

Survey of the Refrigeration Field as it Exists Today, A 
Hosmer, Helen R. 

Electro-culture. A Resume of the Literature 

Translation: The Theory of Lubrication. L. Ubbe- 

lohde 966. 1074. 

Hoxie. C. A. 

Ten-to-one Ratio for Comparing Precision Resistance 

Standards. A 915 

Jackson. C. 

Fort Wayne Electric Rock Drill. The 273 

on, Jas. A. 

Possibilities Open to the Centra! Station in Solving 
the Freight Terminal Problem, The . . 1142 

Johnson. E. S. 

for the Chicago. Mil- 
waukee & St. Paul Railway Company. The 1154 

High-voltage Direct-current Substation Machinery.. 641 
Johnson. H. R. 

Current Supply for Motion Picture Machines 895 

Subdivision of Power as Solved by the Small Motor. 

The. 555 

B. W. 
Short Method for Calculating the Starting Resistance 

for Shunt. Induction and Series Motors. A 131 

Kimball. Fred M. 

Electricity in the Automobile Industry . 550 

Kimball. E. E. 

haracteristic Curves 296 

of the Bethlehem-Chile Iron Mines 






Langrnuir. Irving 

Pure Electron Discharge and its Application in 

Radio-telegraphy and Telephony, The 327 

Larson. C. W. 

Modern Mine Haulage Motor. The 264 

Towing Locomotives for the Panama Canal, The. . . . 101 
Layng. J. F. 

Economies in Operating Small Cars 790 

Selection of Railway Equipment. The 126 

Relation between Car Operation and Power Consump- 
tion 973 

Lehn, H. C. 

Parallel Operation of Alternating Current Generators 

Driven by Internal Combustion Engines 167 

Lilienthal, Jesse W. 

Welfare Work 1092 

Lincoln, Paul M. 

Trend of Electrical Development. The 784 

Liston. John 

Developments in Electrical Apparatus During 1914 . . 80 
Electrical Equipment of the Vermont Marble Company 1015 
Lyle. F. W. 

Growth of Current in Circuits of Negative Tempera- 
ture Coefficient of Resistance 1129 

Mackall. K. W. 

Sign and Building Exterior Illumination by Projection 282 
Mahan. H. E. 

Eye and Illumination, The 268 

Mann. Arthur S. 

Some Problems in Burning Powdered Coal 920. 959 

Mason. L. A. 

Lock Entrance Caissom for the Panama Canal 210 

Minton. J. P. 

Electrophysics: Cathode Rays and their Properties. . 118 
Electrophysics: Electron Theory of Electric Conduc- 
tion in Metals 204 

Electrophysics: Application of the Electron Theory 

to Various Phenomena 287 

Electrophysics: Electromagnetic Radiation from the 

Viewpoint of the Electron Theory 387 

Electrophysics: Some Characteristics of Cathode Ray 

Tubes 636 

Moody, W. S. 

Principal Factors Governing the Choice of Method of 
Cooling Power Transformers as Related to their 

First Cost and Operating Conditions 839 

Moore, C.X. 

X-ray Examination of Built-up Mica 195 

Morrison, G. F. 

Electric Lamp Industry. The 497 

Review of the N.E.L.A. Lamp Committee Report... 925 
Moss. Sanford A. 

Test for Dirt in an Air Supply 622 

Muir, R. C. 

45.000-kw. Synchronous Converter Substation of the 
Aluminum Company of America at Massena Springs. 

N. Yi. The 873 

Xixdorff. S. P. 

Xotes on the Operation of Transformers used with 2 

kw., 100.000 Cycle Alternator 308 

Parham. E. C. 

Practical Experience in the Operation of Electrical Ma- 
chinery, 56, 146. 21 7, 304. 401 , 666. 861 . 928. 1003. 1082. 1 1 4H 
Peck, E. P. 

High Voltage Arrester for Telephone Lines 189 

Peek, F. W„ Jr., 

Electrical Characteristics of Solid Insulations 1050 

High Frequency 934 

Law of Corona and Spark-over in Oil, The 821 

Effect of Altitude on the Spark-over Voltages of Bush- 
ings. Leads, ar.d Insulators ." 137 

Porter. L. C. 

Incandescent Lamps for Projectors 371 

Lighting of Ships, The 143 


Porter, L. C. 

Portable Searchlights for Fire Departments 1144 

Sign and Building Exterior Illumination by Projection 282 
Poor. F. H. 

Ball Bearings in Electric Motors 631 

Potter, W. B. 

Review of Electric Railways, A 444 

Pratt, W. H. 

Some Notes on Induction Meter Design 277 

Randolph, C. P. 

Electric Heating and Heating Appliances 523 

Rea, N. L. 

Water Rheostats 1001 

Reid, Robert 

Process of Impregnating Coils, and a Large Modern 

Impregnating Plant, The 48 

Rettew, G. H. 

Parallel Operation of Frequency Changers 836 

Proper Construction of Earth Connections 904 

Rice. Jr. E. W. 

Status of the Engineer, The 234 

Rohrer, A. L. 

Series of Electrical Tests made in 1883 and their 

Influence on Modern Testing. A 22 

Rohrer, C. J. 

Electricity in Agriculture 483 

Ross, A. A. 

Operating Conditions of Railway Motor Gears and 

Pinions 249 

Roux, George P. 

Emergency Transformer Connections 832 

Open-Delta or V Connection of Transformers 52 

Ruder, W. E. 

Effect of Chemical Composition Upon the Magnetic 

Properties of Steel, The 197 

Rushmore, D. B. 

Electric Power Industry. The 427 

Use of Electricity in Mining Work, The 527 

Ryan, W. D'A. 

Illumination of the Panama-Pacific International 

Exposition 579 

Schildhauer, Edward 

Electricity in the Construction and Operation of the 

Panama Canal (Supplement to July Review) 679 

Shafer, Don Cameron 

"Home Electrical" at the Panama-Pacific Interna- 
tional Exposition, The 572 

Shanklin, G. B. 

Insulation Testing 1008 

Shelton, E. K. 

Protection of Railway Signal Circuits Against 

Lightning Disturbances 1127 

Simmons, L. A. 

Standard in Refrigeration. A 1171 

Slichter, W. I. 

New Advanced Course in Electrical Engineering at 

Columbia University, The 940 


Steinmetz, C. P. 

Control and Protection of Electric Systems 887 

Individual and Corporate Development of Industry, 

The 813 

Protection and Control of Industrial Electric Power. . 979 
Stickney, G. H. 

Factory Lighting 67 

High Candle-power Mazda Lamps for Steel Mill 

Lighting 377 

Stone, C. W. 

Brief Review of the Electric Lighting Industry. A... 439 
Spring, H. E. 

Tests of Large Steam Hoists 179 

Thirlwall. J. C. 

Jitney Problem, The 604 

Power Consumption of Railway Motors 944 

Thomas. 0. E. 

Osbcrne Electriquette. The 299 

Thomson, Stuart 

Methods of Obtaining High Potential Direct Current 1084 
Thomson, Elihu 

Wireless Transmission of Energy 316 

Tinson, H. A. 

Modern Street Lighting with Mazda Lamps 659 

Torrens, J. H. 

Hydro-electric Installation on a Coffee Plantation, A. 219 
Tressler, M. E. 

Cathode Ray Tube and Its Application, The 816 

Trumbull, Horace Niles 

Modern Acid-Dipping, Electroplating and Japanning 

Plant, A 1121 

Ubbelohde, L. 

Theory of Lubrication, The 966. 1074. 1118 

Weaver, S. H. 

Mechanical Effects of Electrical Short Circuits 1066 

Weed, J. Murray 

Mechanical Stresses in Shell Type Transformers. ... 60 
Theory of Electric Waves in Transmission Lines. . . . 1148 

White, W. C. 

Radiotelephony 38 

Whitney, W. R., Dr. 

Research 1012 

Relation of Research to the Progress of Manu- 
facturing Industries, The 868 

Wolf. H. M. 

General Notes on Groundin ; 991 

Woodward, Wm. P. 

High Potential Transformer Testing Equipment 398 

Yates, W. C. 

Electric Motor in the Printing Industry, The 1136 

Yensen, Trygve D. 

Iron-cobalt Alloy, FejCo, and its Magnetic Prop- 
erties, The 881 

General Electric Review 


,, „ t»t^t- T7J* taum r> Ti-cTtT-n'r*'!* Associate Editor, B. M. EOFF 

Manager. M. P. RICE Ed.tor. JOHN R. HEWETT ^.^ ^.^ & & SANDERS 

Subscription Rates: United States and Mexico, $2.00 per year; Canada, $2.25 per year; Foreign, $2.50 per year; payable in 
advance. Remit by post-office or express money orders, bank checks or drafts, made payable to the General Electric Review, 
Schenectady, N. Y. 

Entered as second-class matter, March 26, 1912, at the post-office at Schenectady, N. Y., under the Act of March, 1879. 

VOL. XVIII., NO. 1 h y Ge.^E&uiUany JANUARY, 1915 



Frontispiece 2 

Editorial : The Paths of Progress 3 

The Electrification of the Puget Sound Lines of the Chicago, Milwaukee & St. Paul Railway 5 

By A. H. Armstrong 
The 1500- Volt Direct-Current Electrification of the Ontario Municipal Railway . 10 

By G. H. Hill 
The 2400- Volt Railway of the Bethlehem-Chile Iron Mines Company ... .12 

By E. E. Kimball 

Electro-Culture, a Resume of the Literature 14 

By Helen R. Hosmer 
A Series of Electrical Tests Made in 1883 and Their Influence on Modern Testing . . 22 

By A. L. Rohrer 

An X-Ray Inspection of a Steel Casting 25 

By Dr. Wheeler P. Davey 

A Transmission Line Calculator 28 

By Robert W. Adams 

Some Notes on Magnetization Curves .31 

By John D. Ball 
The Apprentice System at the Lynn Works of the General Electric Company ... 35 

By Theodore Bodde 

Radiotelephonv 38 

By W. C. White 

The Supplying of Power to the Quaker Oats Company 42 

By J. M. Drabelle 

Semi-Outdoor Portable Substation for Berkshire Street Railway 44 

By W. D. Bearce 
The Process of Impregnating Coils; and a Large, Modern Impregnating Plant ... 48 

By Robert Reid 

Open-Delta or V-Connection of Transformers 52 

By George P. Roux 

Practical Experience in the Operation of Electrical Machinery 56 

Loose Commutator; Loose Connection; Low Power-Factor; Hot'Box Indications; Burn- 
out Due to Core Loss; Alternator Speed Low; Loose Core; Loose/Belts; Erratic Elevator 
Speed; Brush-holders Shifted. 

By E. C. Parham 
Mechanical Stresses in Shell-Type Transformers .... .... 60 

By J. Murray Weed 
A Survey of the Refrigeration Field as it Exists Today . .... 65 

By H. I. Holleman 

Factory Lighting .... 67 

By G. H. Stickney 

Notes on the Activities of the A- I- E. E. . 71 

From the Consulting Engineering Department of the General Electric Company . . 73 

Question and Answer Section .74 

In Memoriam: Douglas S. Martin 76 



3 u 

a xi 

>> a 

- u 

OS 2 
(X, u 

•a t; 





It is particularly gratifying that we are 
able to announce the closing of so large and 
important a contract for steam road electri- 
fication as that of the Chicago, Milwaukee 
& St. Paul Railway in this the first issue of a 
new year. We have included in this issue 
also a brief description of the 1500-volt 
direct-current electrification of the Ontario 
Municipal Electric Railways, and the 2400- 
volt railway of the Bethlehem-Chile Iron 
Mines Company. We feel that the very fact 
that such important work as the above under- 
takings represent is being actively pushed at 
the present time should be a distinct en- 
couragement, as showing a marked improve- 
ment in the industrial and financial conditions 
and a faith in the immediate future of the 
economic status of the country. 

The electrification of the Puget Sound Lines 
of the Chicago, Milwaukee & St. Paul Rail- 
ways is the most important steam road elec- 
trification ever undertaken or even seriously 
contemplated; in fact the letting of this 
particular contract would seem to mark a 
new era in electric railway work. The initial 
work includes the electrification of one com- 
plete engine division 1 13 miles in length 
and the total mileage, when yards, sidings, 
etc., are considered, amounts to IBS miles. 
This work is already under way and in the 
early future, if the initial work proves success- 
ful, three additional engine divisions will be 
electrified, making approximately 440 miles 
of main line track or a total of 650 miles, when 
yards, sidings, etc., are included. It would 
appear that all this work is well assured and 
plans are even being made to extend the 
electrified zones to the coast which would 
mean S50 route miles of main line steam road 
converted to electric operation. 

One of the most interesting, and at the 
same time most important features concern- 
ing this large contract is that the change in 
motive power is not being brought about by 
any local conditions such as the necessity 

of abating the smoke nuisance, but is being 
made by the railway company on the straight 
plea of the economies that are to be secured 
by electric traction. The operating results 
of the Butte, Anaconda & Pacific Railway, 
which we published in the November issue 
of the Review, would indicate that there is 
every justification for anticipating economies 
that will more than offset the added interest 
charges on the capital to be expended in 
effecting the change. 

The whole engineering world that is in- 
terested in railway work will undoubtedly 
pay special attention to the fact that the 
three contracts we have mentioned in this 
editorial are all to be operated at higher 
direct-current potentials. The Chicago, Mil- 
waukee & St. Paul Railway will operate at 
3000 volts, the Ontario Municipal Railways 
at 1500 volts, and the Bethlehem-Chile Iron 
Mines Railway at 2400 volts. It surely 
must be considered a most significant fact 
that, in such a very great percentage of the 
large contracts that have been placed during 
recent years in this country for heavy trac- 
tion work, higher direct-current potentials 
have been specified. The reason for this is 
undoubtedly the success that has already 
been achieved with direct-current apparatus 
working at higher voltages. The very fact 
that the Chicago, Milwaukee & St. Paul Rail- 
way Company has adopted a trolley potential 
of 3000 volts shows that the limit had not 
previously been reached where economies 
could be secured by increasing the trolley 
potential without sacrificing any of the vital 
attributes of traction work, such as safety, 
reliability of operation and an all-round effi- 

Another point of great interest concerning 
two of these electrifications, namely, the 
Chicago, Milwaukee & St. Paul and the rail- 
way of the Bethlehem-Chile Iron Mines 
Company is that the locomotives are to be 
provided with regenerative control. On an 
electric railroad scheme of the magnitude 


of the Chicago. Milwaukee & St. Paul Rail- 
way distinct operating advantages should 
result from the provision of electric braking 
for the heavy trains on the steep down 
grades, necessarily encountered in railroad 
work in such mountainous regions. This is 
distinctly in line with the "safety first" 
policy of modern railroading, as electric 
braking removes any danger of accident due 
to overheated brakeshoes and wheels, and 
furthermore results in pow T er economy and a 
lower cost of maintenance. 

The direct-current railway motor has long 
been recognized as the most reliable, efficient 
and flexible means of delivering power to the 
drivers of a locomotive and now that direct- 
current regenerative braking has become an 
accomplished fact it makes the high voltage 
direct-current system most admirably fitted 
to fulfill all the requirements of general steam 
railroad electrification. We consider the 
introduction of electric braking, while still 
retaining the well tried and proved direct- 
current apparatus, to be a distinct step in 
the advance of the art. 

Referring to the Chicago, Milwaukee & St. 
Paul electrification, each locomotive, of 260 
tons local weight, will have 200 tons on 
drivers, an equipment of eight motors, having 
a combined rating of 3440 horse power, and 
a hauling capacity of 2500 tons trailing load 
on a one per cent grade at a speed of approxi- 
mately 16 miles per hour. This great hauling 
capacity, combined with such a high speed on 
ruling grades as 16 miles per hour, is of par- 
ticular interest to the steam railway operator 
who has been educated in the school of Mallet 
operation, in which speeds as low as seven 
miles per hour constitute frequent practice. 
The introduction of such an advanced type 
of motive power should result in somewhat 
radical changes in the methods of operation 

standardized with the use of the steam 

The adoption of a trolley potential of 3000 
volts enables an economic distribution of the 
feeder copper with the spacing of substations 
35 miles apart. The reduction of the neces- 
sary substation apparatus that will be secured 
in this manner, in spite of the fact that such 
heavy trains are to be hauled up mountain 
grades, brings the cost per mile of track 
electrified down to a very reasonable figure, 
and further, it emphasizes the sturdy capa- 
bilities of the direct-current substation ap- 

Mr. A. H. Armstrong in his article shows 
the ample provisions that have been made 
for power supply, and that owing to favorable 
local conditions the railway company has 
been enabled to enter into a contract whereby 
energy will be supplied at 0.536 cents per 
kw-hr. based on a 60 per cent load factor. 
Such figures for energy, even when taken in 
bulk, are unusual and can only be obtained 
in cases where the hydro-electric resources 
have been so wisely conserved and so thor- 
oughly developed as in the case of the 
Montana Power Company. If such thorough 
developments take place in other localities it 
will play an important part in stimulating the 
futher electrification of our steam railways. 

As we said at the outset, we hope that work 
of such a magnitude as that described in this 
issue being undertaken at this time will 
encourage others to look on the bright side 
of present conditions. As this is the first 
issue of a new year, it seems appropriate to 
express the hope that 1915 may be full of 
prosperity and that we may have the pleasure 
of recording many notable steps of progress 
in the engineering arts and industrial re- 
search during the next twelve months in this 


By A. H. Armstrong 

Assistant Engineer, Railway and Traction Engineering Department, 
General Electric Company 

The author gives a brief account of the scope of the work to be undertaken on this the most important 
of steam road electrifications. He gives a description of the power supply available, the cost of power to the 
railway company, the type of substation and rolling stock equipment, and the overhead construction to be 
adopted. It is of special interest to note that the trolley potential is to be 3000 volts, which is the highest 
direct-current potential yet adopted in this country for railway work. — Editor. 

Plans for the electrification of the first 
engine division of the Chicago, Milwaukee & 
St. Paul Railway have now been completed 
and contracts have been let with the General 
Electric Company for electric locomotives, 
substation apparatus and line material, and 
with the Montana Power Company for the 
construction of transmission and trolley lines. 
The initial electrification of 113 miles of 
main line between Three Forks and Deer 
Lodge is the first step toward the electrifi- 
cation of four engine divisions extending 
from Harlowton, Montana, to Avery, Idaho, 
a total distance of approximately 440 miles 
with approximately 650 miles of track, 
including yards and sidings. While this 
comprises the extent of track to be 
equipped in the near future, it is understood 
that plans are being made to extend the 
electrification from Harlowton to the Coast, 
a distance of 850 miles, should the operating- 
results of the initial installation prove as 
satisfactory as anticipated. 

The plans of the Chicago, Milwaukee & 
St. Paul Railway are of especial interest, as 
this is the first attempt to install and operate 
electric locomotives on tracks extending over 
several engine divisions, under which condi- 
tions it is claimed the full advantage of 
electrification can be secured. The various 
terminal and tunnel installations made in the 
past have been more or less necessarv by 
reason of local conditions, but the electrifi- 
cation of the Chicago, Milwaukee & St. Paul 
is undertaken purely on economic grounds 
with the expectation that superior operating 
results with electric locomotives will effect 
a sufficient reduction in the present cost of 
steam operation to return an attractive 
percentage on the large investment required. 
If the savings anticipated are realized in the 
electric operation of the Chicago, Milwaukee 
& St. Paul Railway, this initial installation 
will constitute one of the most important 
mile-stones in electric railway progress, and 

it should foreshadow large future develop- 
ments in heavy steam road electrification. 
The success of electric operation on such a 
large scale will at least settle the engineering 
and economic questions involved in making 
such an installation, and will limit the future 
problems of electrification to the ways and 
means of raising the required capital to effect 
the change in motive power. 

The first step taken towards electrification 
by the Chicago, Milwaukee & St. Paul 
Railway was to enter into a contract with 
the Montana Power Company for an adequate 
supply of power over the 440 miles of main 
line considered for immediate electrification. 
The precautions taken both by the Railway 
Company and Power Company to safeguard 
the continuity of power supply should guaran- 
tee a reliable source of power, subject to few 
interruptions of a momentary nature only. 

The Montana Power Company covers a 
great part of Montana and part of Idaho 
with its network of transmission lines which 
are fed from a number of sources of which 
the principal are tabulated below: 

Madison River 11,000 kw. 

Canyon Ferry 

7,500 kw. 

. 14,000 kw. 

Big Hole 

Butte, steam turbine 

Rainbow Falls 

Small powers aggregat 


3,000 kw. 

5,000 kw. 

21,000 kw. 

7,390 kw. 

Total power developed 68,890 kw. 

Further developments part of which are 
under construction are as follows: 

Great Falls 85,000 kw. 

Holter 30,000 kw. 

Thompson Falls 30,000 kw. 

Snake River 20,000 kw. 

Missoula River 10,000 kw. 

Total power undeveloped 175,000 kw. 

Total power capacitv developed and un- 
developed, 244,000 kw." 


The several power sites are interconnected 
bv transmission lines; the earlier ones are 
supported on wooden poles and operate at 
50,000 volts and the later installations are 
supported on steel towers and operate at 
100,000 volts. Ample water storage capacity 
(300,000 acre-feet), is provided in the Hebgen 
Reservoir and this is supplemented by 
auxiliary reservoir capacity at the several 
power sites which brings the total up to 
418,000 acre-feet. The Hebgen Reservoir is 
so located at the head waters of the Madison 
River that water drawn from it can supply 
in turn the several installations on the 
Madison and Missouri rivers, so that the 
same storage water is used a number of 
times, giving an available storage capacity 
considerablv greater than is indicated bv the 

which will permit feeding each substation 
from two directions and from two or more 
sources of power. This transmission line 
will be constructed with wooden poles and 
suspension tvpe insulators, and will operate 
at 100,000 volts. It will follow in general 
the right of way of the Railway Company, 
except where advantage can be taken of a 
shorter route over public domain to avoid 
the necessarily circuitous line of the railwav 
in the mountain districts. 

The immediate electrification of 113 miles 
will include four substations containing step- 
down transformers and motor-generator sets 
with the necessary controlling switchboard 
apparatus to convert 100,000 volts, 60 cycles, 
three-phase power to 3000 volts direct 
current. This is the first direct-current 









r -\ 



















































Miles from St Paul 
Profile of Section of the Chicago, Milwaukee 8b St. Paul undergoing Electrification 

figures given. It would seem, therefore, 
in changing from coal to electricity as a 
source of motive power, that the railroad is 
amply protected as regards reliability and 
continuity of power supply. 

Due to the great facilities available and the 
low cost of construction under the favorable 
conditions existing, the Railway Company 
will purchase power at a contract rate of 
0.536 cents per kilowatt-hour, based upon a 
60 i ter cent load-factor. It is expected under 
these conditions that the cost of power for 
locomotives will be considerably less than 
is now expended for coal. The contract 
between the Railway and Power Companies 
provides that the total electrification between 
Harlowton and Avery, comprising four engine 
will be in operation bv lanuarv 1 


In order to connect the substations with 

the several feeding-in points of the Montana 

ission lines, a tie-in transmission 

line is being built by the Railwav Company 

installation using such a high potential as 
3000 volts, and this system was adopted 
in preference to all others after a careful 
investigation extending over two years. The 
2400-volt direct-current installation of the 
Butte. Anaconda & Pacific Railway in the 
immediate territory of the proposed Chicago. 
Milwaukee & St. Paul electrification has 
furnished an excellent demonstration of 
high-voltage direct-current-locomotive opera- 
tion during the past year and a half, and 
the selection of 3000 volts direct current for 
the Chicago, Milwaukee & St. Paul Railway 
was due in a large measure to the entirely 
satisfactory performance of the Butte, Ana- 
conda & Pacific installation. 

The equipment for this road was also 
furnished by the General Electric Company, 
and a comparison based on six months steam 
and electric operation shows a total net 
saving of more than 20 per cent on the 
investment or total cost of electrification. 
These figures, of course do not take into 


account the increased capacity of the lines, 
improvement to the service, and the more 
regular working hours for the crews. The 
comparison also shows that the tonnage 
per train has been increased by 35 per cent, 
while the number of trains has been decreased 
by 25 per cent, with a saving of 27 per cent 
in the time required per trip. 


The substation sites of the Chicago, 
Milwaukee & St. Paul Railway electrified 
zone provide for an average intervening 
distance of approximately 35 miles, notwith- 
standing that the first installation embraces 

L500-volt direct-current generators connected 
permanently in senes for 3000 volts. The 
fields of both the synchronous motors and 
direct-current generators will be separately 
excited by small generators direct-connected 
to each end of the motor-generator shaft. 
The direct-current generators will be com- 
pound wound and will maintain constant 
potential up to 150 per cent load and will have 
a capacity for momentary overloads up to 
three times their normal rating. To insure 
good commutation on these overloads the 
generators are equipped with commutating 
poles and compensating pole face windings. 
The svnchronous motors will also be utilized 

Map showing Section of the Chicago, Milwaukee & St. Paul to be Electrified 

20.8 miles of two per cent grade westbound 
and 10.4 miles of 1.66 per cent grade east- 
bound over the main range of the Rocky 
Mountains. With this extreme distance 
between substations and considering the 
heavy traffic and small amount of feeder 
copper to be installed, it becomes apparent 
that such a high potential as 3000 volts 
direct current permits of a minimum invest- 
ment in substation apparatus and consider- 
able latitude as to location sites. 

The substations will be of the indoor type, 
the transformers being three-phase, oil-cooled, 
with 100.000-volts primary and 2300 volts 
secondary windings. The synchronous motors 
will operate at the latter potential. The 
transformers will be rated 1900 and 2500 kv-a. 
and will be provided with four 2Y2 per cent 
taps in the primary, and 50 per cent starting 
taps in the secondary. 

The motor-generator sets will comprise a 
(iO-cycle synchronous motor driving two 

as synchronous condensers and it is expected 
that the transmission line voltage can be so 
regulated thereby as to eliminate any effect 
of the fluctuating railway load. 

The location and equipment of the several 
substations is as follows: 

Morel, two 2000-kw. motor-generator sets; 
Janey, three 1500-kw. motor-generator sets; 
"Piedmont, three 1500-kw. motor-generator 
sets; and Eustis. two 2000-kw. motor-gener- 
ator sets. 

Overhead Construction 

Trolley construction will be of the catenary 
type in which a 4/0 trolley wire is flexibly 
suspended from a steel catenary supported 
on wooden poles, the construction being 
"bracket" wherever track alignment will 
permit and "cross span" on the sharper 
curves and in the yards. Steel supports 
instead of wooden poles will be used in yards 
where the number of tracks to be spanned 


exceeds the possibilities of wooden pole 
construction. Poles for the first installation 
are already on the ground and 30 miles of 
poles are set. Work in this direction will be 
pushed with all speed and will be completed 
ready for operation in the fall of 1915, on the 
delivery of the first locomotives. 

As the result of careful investigation and 
experiments a novel construction of trolley 
will be installed composed of the so-called 

will weigh approximately 200 tons and will 
have a continuous capacity greater than any 
steam or electric locomotive yet constructed. 
Perhaps the most interesting part of the 
equipment is the control, which is arranged 
to effect regenerative electric braking on 
down grades. This feature as yet has never 
been accomplished with direct-current motors 
on so large a scale. The general characteristics 
are tabulated below. 

Outline of 260-Ton Electric Locomotive for the Chicago, Milwaukee & St. Paul Railway 

twin-conductor trolley. This comprises two 
4 wires suspended side by side from the 
same catenary by independent hangers alter- 
nately connected to each trolley wire. This 
form of construction permits the collection 
of very heavy currents by reason of the twin 
contact of the pantograph with the two trolley 
wires and also insures sparkless collection 
under the extremes of either heavy current 
at low speed or more moderate current at very 
high speeds. It seems that the twin-conductor 
type of construction is equally adapted to 
the heavy grades, calling for the collection 
of very heavy currents, and on the more level 
portions of the profile where maximum 
speeds of 60 m.p.h. will be reached with the 
passenger trains having a total weight of 
over 1000 tons. The advantage of this type 
of construction is due partly to the greater 
surface for the collection of current, and 
partly to the very great flexibility of the 
alternately suspended trolley wires, a form 
of construction which eliminates any tendency 
to flash at the hangers either at low or high 
speed. Including sidings, passing and yard 
tracks, the 113 miles of route-mileage is 
increased to approximately 168 miles of 
single track to be equipped between Deer 
Lodge and Three Forks in the initial installa- 

The locomotives to be manufactured bv 
the General Electric Company are of special 
interest for many reasons. They are the first 
locomotives to be constructed for railroad 
service with direct-current motors designed 
for so high a potential as 3000 volts. They 

Total weight 

Weight on drivers 

Weight on each guiding truck 

Number of driving axles. . 

Number of motors 

Number of guiding trucks 

Number of axles per guiding truck 

Total length of locomotive 

Rigid wheel-base 

Voltage of locomotive. . . . 

Voltage per motor 

H.P. rating one hour — each motor 

H.P. rating continuous — each motor. . 

H.P. rating one hour — complete loco- 

H.P. rating continuous — complete loco- 

Trailing load capacity, two per cent. . . 

Trailing load capacity, one per cent. . . . 

Approximate speed at these loads, and 

260 tons 

200 tons 

30 tons 





1 12 feet 

10 feet 







1250 tons 

25110 tons 

10 m.p.h. 

The Chicago, Milwaukee & St. Paul Rail- 
way, from Harlowton to the Coast crosses 
four mountain ranges. The Belt Mountains 
at an elevation of 576S feet, the Rocky 
Mountains at an elevation of 6350 feet, the 
Bitter Root Mountains at an elevation of 
4200 feet and the Cascade Mountains at an 
elevation of 3010 feet. The first electrification 
between Three Forks and Deer Lodge calls 
for locomotive operation over 20.8 miles of 
two per cent grade between Piedmont and 
Donald at the crest of the main Rocky 
Mountain Divide, so that the locomotives 
will be fully tested out as to their capacity 
and general sen-ice performance in over- 
coming the natural obstacles of the first 
engine division. The initial contract calls 
for nine freight and three passenger locomo- 


tives having the above characteristics. The 
freight and passenger locomotives are similar 
in all respects except that the passenger 
locomotives will be provided with a gear 
ratio permitting the operation of 800 tons 
trailing passenger trains at approximately 
(50 m.p.h., and will furthermore be equipped 
with an oil-fired steam-heating outfit for the 
trailing cars. The interchangeability of all 
electrical and mechanical parts of the freight 
and passenger electric locomotives is con- 
sidered to be of very great importance from 
the standpoint of operation and main- 

The cab consists of two similar sections 
extending practically the full length of the 
locomotive. Each section is approximately 
52 feet long and the cab roof is about 14 feet 
above the rail exclusive of housings for the 
ventilation. The trolley bases are about five 
feet about the roof owing to the unusual height 
of the trolley wire which will be located at a 
maximum elevation of 25 feet above the 
rail. The outer end of each cab will contain 
a compartment for the engineer while the 
remainder is occupied by the electric control 
equipment, train heater, air-brake appa- 
ratus, etc. 


The eight motors for the complete locomo- 
tive will be Type GE-253-A. This motor 
has a normal one-hour rating of 430 h.p. 
with a continuous rating of 375 h.p. The 
eight motors will thus give the locomotive 
a one-hour rating of 3440 h.p. and a con- 
tinuous rating of 3000 h.p. which makes it 
more powerful than any steam locomotive 
ever built. The tractive effort available for 
starting trains will approximate 120,000 lb. 
at 30 per cent coefficient of adhesion. 

Each motor will be twin-geared to its 
driving axle in the same manner as on the 
Butte, Anaconda & Pacific, the Detroit 
River Tunnel and the Baltimore and Ohio 
locomotives, a pinion being mounted on each 
end of the armature shaft. The motor is of 
the commutating-pole type and has openings 
for forced ventilation from a motor-driven 
blower located in the cab. 

The freight locomotives are designed to 
haul a 2500-ton trailing load on all gradients 
up to one per cent at a speed of approximately 
1G m.p.h., and this same train load, unbroken, 
will be carried over the 1.66 and two per cent 
ruling grades on the west and east slope of the 
Rocky Mountain Divide with the help of a 

second similar freight locomotive acting as 
a pusher. Track provision is being made at 
Donald, the summit of the grade, to enable 
the pusher locomotive to run around the 
train and be coupled to the head end to 
permit electric braking on down grade. In 
this case the entire train will be under 
compression and held back by the two loco- 
motives at this head end, the entire electric 
braking of the two locomotives being under 
the control of the motorman in the operating 
cab of the leading locomotive. It is con- 
sidered that electric braking will prove very 
valuable in this mountain railroading, as in 
addition to providing the greatest safety in 
operation, it also returns a considerable 
amount of energy to the substations and 
transmission system which can be utilized 
by other trains demanding power. In this 
connection, the electric locomotives will have 
electric braking capacity sufficient to hold 
back the entire train on down grades, leaving 
the air-brake equipment, with which they 
are also equipped, to be used only in emer- 
gency and when stopping the train. There is 
therefore provided a duplicate braking system 
on down grades which should result in safety 
of operation, and should eliminate break- 
downs, wheel and track wear and overheating, 
as well as leading to a reduction in mainte- 
nance and an improvement in track condi- 

With the completion of the remaining 
engine divisions it is proposed to take ad- 
vantage of the possibilities afforded by the 
introduction of the electric locomotive by 
combining the present four steam-engine 
divisions into two locomotive divisions of 
approximately 220 miles length; changing 
crews, however, at the present division points. 
As the electric locomotive needs inspection 
only after a run of approximately 2000 miles, 
requires no stops for taking on coal or water, 
or layover due to dumping ashes, cleaning 
boilers or petty roundhouse repairs, it is 
expected that the greater flexibility of the 
locomotive so provided will result in con- 
siderable change in the method of handling 
trains now limited by the restrictions of the 
steam engine. 

The electrification of the Chicago, Mil- 
waukee & St. Paul Railway is under the 
direction of Mr C. A. Goodnow, Assistant 
to the President in charge of construction, 
and the field work is under the charge of 
Mr. R. Beeuwkes, Electrical Engineer of the 




By G. H. Hill 

Assistant Engineer, Railway and Traction Engineering Department, 
General Electric Company 

The general faith in higher direct-current potentials for railway work is exemplified by the equipment 
selected for the important work described by the author. The work now being done on the Ontario Municipal 
Railway is only the nucleus of much more extensive undertakings by the same road in the future. The 
scope of the present work is shown in the accompanying article. — Editor. 

Various municipalities of the Province of 
Ontario, Canada, have been considering for 
some time an extensive system of inter- 
connecting electric railways. The general 
scheme provides for a network of interurban 
roads supported, as it were, by a backbone of 
main line electrification handling a relatively 
heavy freight service. The links connecting 
the cities and towns will be municipally 
owned and operated, thus continuing the 
already extensive municipal ownership fea- 
ture existing in Canada. 

The scheme, in brief, contemplates new 
roads, or electrification of existing roads 
through the central part of that portion of 
the Province lying between Lake Erie and 
the Georgian Bay from Sarnia on the west to 
Toronto and Whitby on the east, passing 
through London, St. Mary's and Guelph. 
This route follows generally the transmission 
system of the Hydro-Electric Power Com- 
mission from Niagara Falls. Several branches 
and connections are contemplated communi- 
cating with all the larger cities in that portion 
of the Province. 

The proximity of the transmission system 
affords excellent facilities for substation 
locations and precludes the necessity of 
building a distributing transmission line 
along a large part of the right-of-way. 

The Hydro-Electric Commission of Ontario, 
Sir Adam Beck, Chairman, and Mr. F. A. 
Gaby, Chief Engineer, will perform the 
double function of consulting engineer and 
contractor for the entire project, installing 
all equipment ready for operation. The 
individual sections of the system will then be 
turned over to ihc municipalities to operate. 

The city of London has taken the initial 
step aid is now proceeding with the elect rifi- 
cation of i hat portion of the main line between 
Port Stanley on Lake Erie and London, about 
2 1 miles north thereof. 

This road has for several years been leased 
by the city of London to the Pere Marquette 

Railroad and forms an important connecting 
link between the Lake Erie freight ferries 
and distributing centers of the Province. The 
chief commodity for transportation is coal 
from Pennsylvania. 

The single track line is approximately 
23.5 miles long and passes through Whites, 
St. Thomas, Glanworth and Westminster, 
connecting with the main lines of the Grand 
Trunk. Michigan Central and Canadian 
Pacific Railroads. The profile includes a 
maximum of 1.0 per cent grade north of 
Port Stanley and contains other grades of 
0.8 per cent and 0.5 per cent for short dis- 

The scheduled service will comprise a 
locomotive freight traffic and multiple-unit 
passenger car trains between the two terminals. 
Sixty-ton locomotives will take loaded freight 
cars from the ferry boats at Port Stanley and 
haul them in trains of approximately 800 
tons to St. Thomas and London, returning 
with trains of empty cars. In addition to the 
above there will be local merchandise freight 
between the stops along the line. The 
passenger service will be performed by 
limited and local trains consisting, a large 
part of the time, of a motor car and one trail 
car providing a half-hourly service in each 
direction, the limited and local alter- 

The Hydro-Electric Commission made a 
careful investigation of operation when using 
both high voltage direct current and single- 
phase alternating current on the trolley, with 
the final adoption of 1500 volts direct current. 

The energy will be supplied from two 
substations: One located at London in an 
extension of the present Hydro-Electric 
substation, and the other at a distance of 
14.2 miles from London near St. Thomas. 
The latter will be a new substation. Each 
will be equipped with' synchronous converters 
with their respective transformers and switch- 
boards, converting from 110,000 volts, 25- 



cycle alternating current to 1500 volts direct 

The overhead structure will be of the 
familiar single catenary type supported on 
side brackets from lattice steel poles placed 
approximately ISO ft. apart on the tangent. 
The 0000 B.&S. copper trolley wire will be 
supplemented by suitable copper feeders. 

The rolling stock covered by the initial 
order placed with the General Electric 
Company includes three 1500-volt, 60-ton 
locomotives, five 4-motor 1500-volt passenger 
car equipments complete with multiple unit 
control and air brakes, and four trail car 
control and air brake equipments. 

The locomotives are of Type 4-0-4 and 
will be carried on two swivel trucks bringing 
all the weight on the drivers, the equipment 
being housed in a steel box type cab extending 
over practically the entire length of the 
locomotive. Each will be provided with four 
GE-251, 750 /1500-volt motors designed for 
750 volts across each armature and insulated 
for 1500 volts. Two motors will be connected 
permanently in series and the two-motor 
groups thus formed will be capable of con- 
nection in series or parallel for speed control. 

The cab will be divided into three compart- 
ments, one at each end for accommodating 
the operator and the intervening compart- 
ment where the control equipment and 
accessories will be located. The operating 
compartments will be provided with 1500-volt 
electric heaters. 

Each of the GE-251 motors will have an 
hourly rating of 245 h.p. with 1500 volts on 
the trolley. At this rating the locomotives 
will exert a tractive effort of 21,500 pounds. 

Control will be effected by a double end 
Type M standard equipment, a master con- 
troller at each operating position actuating 
the main 1500-volt contactors by means of a 
600-volt circuit supplied from a dynamotor. 
Multiple-unit train operation is arranged 
for so that the simultaneous control of three 
locomotives coupled together can be accom- 
plished from any master controller. The 
equipment is also so designed that a locomo- 
tive may haul a train of eight or ten passenger 
trail cars and provide lighting energy for them. 

The current collectors will consist of panto- 
graph slide trolleys having two contact pans 
pressing against the trolley conductor. Two 
of these devices will be furnished on each 

locomotive. They will be electro-pneumati- 
cally controlled from any operating position 
with one, two or three locomotives hauling 
a train. 

Each motor passenger car will be driven 
by four GE-225-750/ 1500-volt fully venti- 
lated commutating-pole motors connected 
two groups of two in series. The one-hour 
rating is 125 horse power with 1500 volts 
on the trolley. 

Each motor car has sufficient capacity to 
haul one trail car and provision is made for 
the motor and trail cars to be operated in 
trains up to a total of three motor and 
three trail cars. All trail cars will be equipped 
with master controllers at each end so that 
multiple-unit train operation is possible 
from either end of any motor or trail 

Control energy for a motor and trailer 
will be derived from a 1500 /600-volt dyna- 
motor on each motor car. The dynamotor 
will also supply energy for lighting one motor 
and one trail car. Alain and auxiliary train 
cables will run continuously throughout a 
train, provision being made for the simul- 
taneous raising and lowering of all panto- 
graphs and also for simultaneous sanding 
(by electro-pneumatic valves) of all cars 
from any operating position. The panto- 
graph trolleys will be identical with those 
on the locomotives. 

Each car will carry a combined straight 
and automatic air-brake outfit of the variable 
release type, with the air supply furnished 
by 1500-volt compressors. The compressor 
governors will all be equalized on a special 
wire running throughout the trains in the 
auxiliary train cable. 

The cars which will be placed in service 
on the London and Port Stanley Railway will 
be built to the Commission's specification. 
They will be all steel, 59 feet long and thor- 
oughly modern in every respect. The motor 
and trail coaches will be identical except 
for motors. The former will weigh approxi- 
mately 51 tons loaded and equipped, while 
the latter will have an approximate loaded 
weight of -')2 tons. 

The growth of the Ontario Municipal 
system will be watched with much interest 
since it marks a rather novel departure on a 
large scale from the usual American pro- 




By E. E. Kimball 

Railway and Traction Engineering Department, General Electric Company 

The work to be done in developing the rich iron mines at Tofo, Chile, includes the construction of a very 
difficult line of railway which is to operate at 2400 volts direct-current, the building of a steam power station 
and transmission line, "the installation of electric shovels and crushers, and the building of a settlement for 
the officers and operators of the company. The author gives a short outline of the work contemplated and 
cites some of the local conditions which make this undertaking one of particular difficulty. — Editor. 

One often notices in technical papers and 
popular magazines articles describing the 
wonderful mineral resources of South America, 
especially large deposits of iron and copper 
ores, but few realize how rapidly and on 
what a large scale these resources are being 
developed and where the products find a 
market. A remarkable deposit of iron ore is 
found at Tofo, Chile, where the Bethlehem- 
Chile Iron Mines Company is preparing to 
mine this ore with the aid of electric power 
and to ship it to the United States for use 
in the blast furnaces at South Bethlehem, Pa. 

These mines occupy the summit of two 
hills, approximately 2000 ft. above sea level 
and about four miles in an air line from the port 
of Cruz Grande. The remarkable feature 
of these mines is that there are great quanti- 
ties of ore in sight and it is nearly pure iron 
(67 per cent Fe.). With the opening of the 
Panama Canal to commerce this ore can 
be mined and shipped by that route from 
Chile to Xew York and thence to South 
Bethlehem, Pa. 

An electric railway operating at 2400 volts 
is now being built to develop these mines. 
In addition to this electric railway the 
development will include a steam-power 
station and high tension transmission lines, 
from the port of Cruz Grande to Tofo. and 
the installation of electric shovels, crushers 
and other machinery for mining operations 
at Tofo. Ore pockets and vessel loading piers 
will be constructed at Cruz Grande. It 
requires also the building of residences for 
officials and establishing ample water supply 
and fire protection as well as the provision 
of an electric lighting system for the villages, 
piers, etc. 

At present a certain tonnage of ore is mined 
by steam drilling and transported to the coast 
over a telpherage system which consists 
of a string of ore buckets suspended from a 
steel cable supported on steel towers and 
operated by the weight of the loaded buckets 
riding, which furnish power for taking 
up the empties. This system is started by a 
it when once started it requires 
no external force to keep it running, in fact. 

part of the energy is dissipated in operating 
a large fan which is used for governing the 
speed. Probably one of the chief reasons for 
adopting this system was on account of the 
small amount of power required to operate 

It was essential that means be provided for 
saving water and fuel, in other words, power. 
Obviously, the power taken by the empty 
trains ascending the grades, if supplied by 
loaded trains descending, would represent a 
saving which could not be effected by minor 
economies of fuel. This feature in the railway 
electrification, therefore, received a great deal 
of attention both from the standpoint of 
saving power and on account of other practical 
operating advantages. 

The railroad from the mines to the piers 
is approximately 15 miles long with an 
average grade for nearly the entire distance 
of three per cent. This is also the maximum 
grade. Its alignment is far from straight, 
as may be seen from the fact that in an air 
line the mines are only four miles from the 
coast, whereas the railroad reaches the same 
height only after traversing 15 miles. 

In the operation of heavy grade sections 
of steam railroads great difficulty has been 
found in getting rid of the heat from brake- 
shoes and wheels, which is another argument 
in favor of electric braking on the locomotives. 
In the study of this problem it was shown that 
regenerative braking could be accomplished 
successfully on a high voltage direct-current 
system, that is, the motors under the loco- 
motive are made to act as generators and 
return energy to the trolley to be used by 
another locomotive ascending, or back to the 
power house where it would help supply the 
demands of the mines. 

These locomotives will weigh 110 tons 
on drivers and will be equipped with four 
300-h.p., 1200 2400-volt motors operated two 
connected permanently in series on 2400 
volts. The initial installation will consist of 
three of these locomotives, each having a 
capacity to haul a 450-ton train up grade at 
10H m.p.h. and exerting the same braking 
effort when regenerating at 12 m.p.h. 



In case the locomotives are operating with 
the maximum train weights down grade a 
portion of the braking will be done with air 
brakes, and when stopping air brakes will be 
used alone. 

The trolley will be of 4/0 grooved copper 
wire, catenary suspended from a steel messen- 
ger supported by a mixture of bracket and 
cross span construction on wood poles. These 
poles will be of cedar and will be shipped 
from the United States, as Chile grows no 
timber suitable for this purpose. A duplicate 
22,000-volt high tension transmission line 
will in general follow the trolley and will be 
carried on the same poles when possible. 
In places, however, it will leave the railroad 
right of way for a more direct route to the 
mines. These transmission lines will supply 
power for the operation of crushers, electric 
shovels, pneumatic tools and machine shops, 
as well as for pumping water and other 
sundry purposes. 

The mining of this ore will be accomplished 
by blasting the ore exactly as in modern 
rock quarries, and then by means of elec- 
tric shovels it will be loaded onto side- 
dump cars, when it will be hauled a short 
distance to the crusher plant and crushed 
to a size suitable for use in blast furnaces. 
The crushed ore will fall by gravity into bins 
ready for loading into hopper cars ; it will then 
be hauled to the vessel loading piers and 
dumped into ore pockets. From here it is 
loaded by gravity into 17,000-ton steel 
vessels specially constructed for this purpose 
and shipped to unloading piers in New Jersey, 
and there loaded onto cars for South Bethle- 
hem, Pa., where it is ready for the blast 
furnaces. The transportation of the ore is, 
therefore, a big item of its cost and every 
facility has been provided to save the expense 
of handling it. 

In the power house oil-fired boilers are to 
be used which will permit of an easy control 
of the heat with every fluctuation of load, 
and because of absence of dirt the usual 
partition between the generator room and 
the boiler room will be omitted so that the 
operators will be able to anticipate changes 
in load in time to make proper adjustments. 
The oil is received in tank vessels and pumped 
to an oil storage tank above the power 
station. From the main reservoir it runs 
by gravity to the auxiliary reservoir near the 
station and is fed to the burners by means 
of a small pump. The boilers will be set high 
so that grates may be installed if there is any 
advantage to be obtained from the use of coal. 

One of the interesting features of this 
installation is the ingenious method employed 
for evaporating boiler "make-up" water 
from sea water. This is done by an evaporat- 
ing condenser through which is "by-passed" 
a part of the exhaust steam from the turbines. 
By adjusting the difference in the vacuum 
between the main and the evaporating 
condensers the amount of evaporation can 
easily be governed without affecting the 
economy of the steam turbines appreci- 

The generating room contains two 3500-kw. 
three-phase, 60-cycle, 2300-volt Curtis G-E 
steam turbines with direct-connected exciters 
for supplying power to the railroad and the 
mines; two 300-kw. three-phase, 60-cycle, 
(i()0-volt turbines for operating motor-driven 
auxiliaries, fire pumps, etc., and at night, 
lights for the piers, villages and mines when 
the main turbines are shut down. To accom- 
plish this result there is installed a small bank 
of step-up transformers so that the high 
tension lines may be energized at minimum 
loss. This arrangement avoids considerable 
complication in the switchboard wiring and 
avoids providing steam-driven auxiliaries 
with complicated steam and water piping. 

Power for the operation of the mines and 
crusher plant is stepped up by means of two 
banks of transformers, each bank consisting of 
three 667-kv-a., 60-cycle, 22, 000 /2300-volt 
oil-cooled transformers. At the mines 
the voltage is stepped down again to 2300 
volts for local distribution. Power for the 
operation of the railroad is taken through two 
1000-kw., three-unit motor-generator sets, 
each consisting of one 1400-kv-a., O.S-p-f. 
three-phase, 60-cycle, 2300-volt synchronous 
motor-generator set direct-connected to two 
500-kw., 1200/2400-volt direct-current gener- 
ators, designed to operate two in series on 
2400 volts. These sets have direct-connected 
exciters on each end for exciting the synchro- 
nous motor and d-c. generators. Space has 
been left in the design of the building for 
future boilers, main turbines and motor- 
generator sets when required. 

The power station building will be located 
on solid rock foundations and a type of 
construction employed which is particularly 
adapted to resist earthquake shocks, which 
are frequent and sometimes violent and 
followed by tidal waves. It will, therefore, 
be located back from the water's edge, on 
high land reasonably safe from these dis- 
turbances. A pump house of very sturdy 
construction will be erected at the water's 



edge to supply circulating water to the 
condensers and for fire protection for the 
village and piers. 

Nearly all the sources of fresh water supply 
are located in the valley behind the mines and 
it requires pumping to the mine level for use 
at the mines. The excess is stored in reser- 
voirs between the mines and Cruz Grande 

for such purposes for which it is suitable and 
also to relieve the pressure on the pipes at 
Cruz Grande. At the present time a great 
deal of this work has been completed and a 
temporary pumping and lighting plant is 
installed near one of these springs and 
supplies water for the mines and lights for 
the villages. 


By Helex R. Hosmer 

Research Laboratory, General Electric Company 

The literature published on electro-culture is extensive but scattered. The writer has made an excellent 
review of this published matter, dividing the subject according to the different methods used to stimulate 
plant life by electricity. The progress, or lack of progress, made by the application of each method, is dis- 
cussed, and the conclusion is reached that these investigations, on the whole, have been too cursory. A valu- 
able list of references is also given. — Editor. 

The scientific literature of the last ten 
vears has contained frequent references to 
the art of increasing plant growth and yield 
bv the application of electric stimuli of cer- 
tain kinds, an art most commonly designated 
as electro-culture. The material given, how- 
ever, represents very little experimental 
work in proportion to its volume, consisting 
in the main of more or less complete his- 
torical review? concluded by a few para- 
graphs describing some recent investigation. 
The effect upon a reader desiring to become 
acquainted with the work done within a 
reasonable length of time is irritating, to say 
the least. In view of the growing interest 
in intensive methods of agriculture, and also 
in methods of filling in the valleys in the load 
curves of central stations, there is reason to 
expect a much more exhaustive investigation 
of this subject in the not remote future. For 
this reason it has seemed desirable to collect 
the facts from the scattered sources, and 
attempt to arrange them in a form more 
convenient for use, that is, from the point 
of view of the invader of the province rather 
than the historian. 

It has been found that the experiments of 
the past fall naturally into five classes, 
differing principally in the method of appli- 
cation ( if elecl rical energy. These methods are : 
Illumination by electric light. 

action of atmospheric electricity 
from an elevated collector to an 
electrode in the soil, or to dis- 
charge points above the plants. 
Constituting the soil the electrolyte 
of a voltaic cell by burying in it 
two plates of dissimilar metal con- 
nected bv a conductor. 





Passing current from an external 
source through the soil between 
electrodes buried therein. 
Production of a silent or glow dis- 
charge through the air from over- 
head antennae to the soil. 
These methods will be taken up in the 
order given, which is approximately that of 
their importance. 


Illumination by electric light. 

There seems to have been relatively little 
work done upon the effect of illuminating 
plants by artificial or electric light. In 1861 
"Herve" Mangon found that electric light 
influences the formation of chlorophyl in a 
way similar to that of sunlight. That the 
absorption and assimilation of carbon dioxide 
occurred as usual under the electric arc was 
shown by Prellieux eight years later. 1 

In 1880 Wilhelm Siemens confirmed these 
observations, but found that under certain 
conditions injurious effects were obtained, 
and hence he used an opalescent glass shade 
over the light. 

These facts were further confirmed by 
Schraier in 1881, and by Bailey, Cornell 
University, in 1891. Bonnier in 1S92, and 
Couchet in 1901, studied the structure 
alteration in plants and the leaf growth in 
relation to the electric light. 

Since 1891 this line of attack has been 
neglected, probably because of the attention 
attracted by the work of Lemstrom, and the 
success of his method. 

Dorsey, however, in 1914, mentions the 
treatment of hothouse radishes and lettuce 
for three hours each day beginning at sunset, 



with red light from a 100-watt lamp, and with 
blue light from a Cooper-Hewitt lamp. The 
lettuce was affected favorably, the radishes 
unfavorably. 2 


Conduction of atmospheric electricity from 
an elevated collector to an electrode in the soil, 
or to discharge points above the plants. 

Among the earliest attempts to apply 
atmospheric electricity to plant culture ap- 
pears to have been that of Abbe Bertholon, in 
1783. He called his apparatus the electro- 
vegetometer. It consisted of a number of 
metal points similar to a lightning rod, sup- 
ported at a considerable elevation, and con- 
nected by a conductor to an iron bar furnished 
with discharge points which hung down just 
over the plants treated. The whole apparatus 
was insulated by wooden supports. The 
Abbe stated that the use of this arrangement 
always produced an increase in the fertility, 
vigor, and growth of the plants. 3 

Later, 1879, Grandeau and his pupil 
LeClerc showed by careful comparative 
measurements, analyses, etc., that protection 
of plants from atmospheric electricity by 
enclosure in wire cages often retards the 
growth over 50 per cent. But Naudlin re- 
peated his experiment a little later with 
results diametrically opposite. The more 
recent experience of Pinot de Moira agrees 
with that of Grandeau. 

A modification of Bertholon's method 
called the geomagnetifere system has been 
quite commonly used in France. This con- 
sists of an elevated conductor connected to 
wires running through the soil under the 
plants to be influenced. A typical installa- 
tion is that of Pinot de Moira at Clifton, 
Eng., which was in operation for several 
years to good advantage. 

Berthelot carried on considerable work at 
Meudon in France. He found that the 
growth of plants on the top of a 28-meter 
tower was greater than at the foot. 

Lieutenant Basty experimented with metal 
rods terminating in a ball of non-oxidizable 
metal at the lower end which was buried in 
the ground as deeply as the roots of the 
plant were likely to penetrate and projected 
from two and one-half feet to six and one- 
half feet above the surface, depending upon 
the plant treated. The first height was used 
for strawberries. He claimed that beneficial 
results were noted about each rod for a radius 
equal to half the height. 4 


Constituting the soil the electrolyte of a 
voltaic cell by burying in it two plates of dis- 
similar metals connected by a conductor. 

Speschenew in Russia obtained marked 
results from plates of different metals buried 
in the ground connected by wire. 

More recently, 1906, Rawson and Le 
Baron 5 have used the same method in green- 
houses. Plates of copper and zinc were sunk 
at opposite ends of lettuce beds and gave a 
potential difference of 0.5 volts and current 
of from 0.4 to 15 milliamperes. The lettuce 
thus treated was ready for market a week 
sooner than that not treated. 

Priestly 3 tried the method of Speschnew, 
using plates of copper and zinc between which 
beans were planted. The plants treated 
appeared two days earlier, developed more 
rapidly, and the average size and weight of 
the mature beans was about a third greater. 
Some other qualitative experiments were 
inconclusive. The current in very damp soil 
was twelve milliamperes between plates of 
200 sq. in., four feet apart. 

Newman, 6 however, states that the results 
of a dozen experiments indicated no effect 
whatever, and that the reports of others have 
been in confirmation of this fact. 


Passing current from an external source 
through the soil between electrodes buried 

This method of plant stimulation has been 
the source of numerous conflicting reports, 
and its applicability seems still to be in doubt. 
A number of investigators have found that 
it increases the rate and proportion of 

E. H. Cook 7 states that this is the only 
effect that he was certain was produced by 
currents of 100 milliamperes at 20 volts. 

Kinney 8 in 1898 and Ahlfvengren in 1899, 
confirmed his results. The former con- 
sidered three volts the optimum, but the 
latter believed this to vary for different 
plants, and, under different conditions, for 
the same plant. Lewenherz's conclusions 
also agreed with the above, but he con- 
sidered also that the direction in which the 
current traversed the seed was of importance. 

Kovessi, 9 1912, on the other hand, as a 
result of over 1100 pot tests, came to the 
conclusion that direct currents through the 
soil are without exception harmful both to 



germination and later growth. Schnecken- 
berg, 10 commenting upon this paper, remarks 
that he ought to have known this fact from 
a knowledge of the simple laws of electro- 
chemistry and endosmosis before performing 
the 1100 experiments, but goes on to point 
out that Kovessi's statement should read 
"horizontal direct currents through the 
soil ' ' and must not be extended to cover any 
other type of electrical treatment. Kovessi 
does not state what strength of current he 

Geriach and Erlwein, 11 1910, describe 
experiments with low potential direct cur- 
rent, 6 volts, 0.2 to 0.4 amperes, at Bromberg, 
upon an area of 914 sq. ft. planted to barley 
and cabbages. Iron plates buried in the soil 
were used for electrodes. The treatment was 
continuous night and day until harvest. 
No beneficial effect was obtained. 

Peaslee, 12 1910, using direct current in 
greenhouse experiments on the germination 
and rate of growth of seedlings, such as 
cauliflower, cabbage, beets, etc., experienced 
failure until he lowered his current density 
and adopted carbon electrodes, which, unlike 
some metals, do not react with the soil to 
form deleterious salts. He obtained the most 
favorable results at a power consumption of 
between 0.5 and 0.(i watts per cu. ft. which 
gave increased fertility of seed, more rapid 
and vigorous development, and increased 
size of plant, especially of the root. In the 
case of a cauliflower, the advantage in respect 
to growth was nearly 150 per cent. Radishes 
carried through to a marketable size had a 
root growth 403 per cent, and a top growth 
1 1 7 per cent greater than the control* plants. 

Similar tests with alternating current were 
consistently negative again until the watts 
per ru. ft. were reduced to 0.0114 (current = 
0.000034 amp. per sq. in.) when an increased 
fertility of 50 per cent and an increased 
growth of 32 per cent was obtained. 

Dorsey, 2 1913, tried some greenhouse 
experiments using direct current (1.5 volts 
arid 0.0003 to 0.07 amp., and 3 to 8 volts and 
ooiio, to 0.05 amp.) and also 60-cycle after- 
current, 110 and 220 volts between 
carbon electrodes. The results were bad in 
The temperature of the treated 
was a degree higher than the controls. 

It is evident that the investigation of this 

lectric treatment has been entirely 

insufficient to lead to any trustworthy con- 

* The expressions "control." "control plants." etc., are used 
comparative experiments carried on 
v- under the same conditions, but without elec- 
trical stimulation. 

elusions. The controlling factors have scarcely 
been indicated as yet. 


Production of a silent or glow discharge 
through the air from overhead antennae to the 

The stimulation of crops by a discharge 
of electricity through the air to the soil seems 
to be the method best founded upon theory 
and most promising in practice. 

Prof. Lemstrom 13 of Helsingfors Univer- 
sity, Finland, first remarked upon the fact 
that the extraordinarily rapid and fruitful 
growth of such vegetation as survives the 
frosts in the Arctic and sub-Arctic regions 
can not be accounted for, as has been sug- 
gested by the long hours of daylight. He 
showed that the total light and heat supplied 
are actually less than at the latitude of 
Petrograd or Christiana, on account of the 
low elevation of the sun above the horizon. 
It has been proved beyond doubt that there 
exist in the atmosphere of these high lati- 
tudes much stronger currents passing to the 
earth than is the case further South. These 
are evidenced by their luminescent effects, 
such as the Aurora. A great proportion of 
the vegetation, especially that peculiar to 
northern regions, is equipped with pointed 
leaves, etc., which are especially adapted 
to electrical discharge. Moreover, in study- 
ing sections of fir trees, Lemstrom found a 
periodicity in the occurrence of especially 
large growth which is the same as that of the 
occurrence of sun spots and auroras, i.e., 
every 10 or 11 years. Lemstrom goes into 
these indications and the further electrical 
phenomena and facts from which he was led 
to begin his investigations in considerable 
detail in his book "Electricity in Agriculture." 

He supected that the electrical influence 
played a part hitherto overlooked in the 
growth of vegetation in other parts of the 
world. With this in view, he tried to redupli- 
cate the conditions of the Arctic by producing 
a similar electrical tension in the atmosphere. 
He applied a positive potential from an in- 
fluence machine of which the negative was 
grounded to a wire network suspended above 
the plants, producing a silent discharge to 
the earth. He worked first with pots, and 
then in the open field. His power consump- 
tion was low, as indicated by the fact that a 
0.1 horse power motor served to drive his 
influence machine. 

Lemstrom extended his researches to dif- 
ferent farms in Finland and, in later years, to 



other countries. The procedure was tested 
under his supervision at Durham College, 
England; in Burgundy, near Breslau in Ger- 
many, and at Atvadaberg. His book con- 
tains full details as to the extent, circum- 
stances, and results of all these experiments. 
As a result of his experience he concludes 
that the minimum increase in yield for all 
crops under the proper conditions should be 
about 45 per cent. For certain crops it may 
rise as high as 100 per cent. Improvement 
occurs whether the network be charged 
positively or negatively to the soil, but 
better results were obtained in the former 
case. The effect is not apparent alone in the 
quantity, but an improvement of quality, 
and a shortening of the period of growth, 
sometimes by 50 per cent, is general. An- 
alyses are given to indicate that in the case 
of grain there is an increase in the proteid 
content. Frequent instances were encoun- 
tered where the electric current was delete- 
rious, but repetition of the work under im- 
proved conditions eliminated the trouble. 
Thus it was found that during drought or 
in hot sunshine the plants suffer harm, and 
certain species, among which may be men- 
tioned peas, cabbages, and carrots, are par- 
ticularly sensitive. Watering, and discon- 
tinuing the treatment during the middle of 
the day has produced equally good results 
with these vegetables. Lemstrom points out 
that lack of uniformity in cultivation, nature 
of soil, and fertilization between the experi- 
mental and control 'plots often leads to 
erroneous conclusions. The better culti- 
vated and fertilized a field is, the larger the 
percentage increase in yield due to electro- 

Lemstrom devotes a chapter to directions 
for the choice and installation of apparatus, 
with an estimate of costs. 

Lemstrom 's procedure suffered from a 
great disadvantage. His influence machine 
was quite inadequate for the purpose, hence 
his overhead wires could not be hung more 
than 16 in. above the plants, which inter- 
fered with economical cultivation of the soil. 

At Gloucester, 3 experiments with a some- 
what more powerful machine, enabling the 
elevation of the wires to five feet above the 
ground gave results with various crops as 
follows : 

Beets, 33 per cent increase. 
Carrots, 50 per cent increase. 
Turnips, increase: not quantitatively meas- 

The beets raised under electrification gave 
on analysis about 14 per cent more sugar 
than the control crop. 

This increase in sugar content has been 
confirmed by almost every investigator, 
irrespective of whether his results were 
favorable to the process in other ways. 

In 1904, Newman 3 , 6 performed some sim- 
ilar tests with a small Wimshurst machine 
driven by an oil engine, operating upon 
fifteen greenhouses, and upon an area in 
the open amounting to about a thousand 
square yards, including control plots. The 
wires were strung about 16 in. above the plant 
tops and were furnished with downward 
directed points of fine wire for discharge 
points. Ordinary telegraph insulators were 
sufficient except in wet weather, when 
almost all the energy was lost by leakage 
down the supports. 

The treatment was applied for a period of 
108 days, 9.3 hours daily, the first half of the 
time mainly by day, the last half by night. 
The results from the electrified plants were 
as follows: 

Cucumbers, 17 per cent increase. 

Strawberries, five-year plants, 36 per cent 

Strawberries, one-year plants, 80 per cent 
increase, and produced more runners. 

Broad beans, 15 per cent decrease, ripened 
five days sooner. 

Cabbages, (spring) mature 10 days sooner. 

Celery, two per cent increase. 

Tomatoes, no effect. 

The cucumbers were all affected by a 
bacterial disease about the middle of their 
growth, and this made much greater headway 
on the non-electrified plants. Aside from 
the troubles with the influence machine and 
oil engine, which were rather inadequate, -the 
installation required no attention except for 
the clearing away of cobwebs and stray 
shoots, etc., from the network. 

This work was continued on a larger scale, 
Newman 3 working in conjunction with Sir 
Oliver Lodge. The latter overcame several 
of the inherent difficulties of the process by 
the invention of a mercury arc rectifier 
supplying a 100,000-volt direct current. The 
new installation consisted of an oil engine 
and dynamo producing three amperes, at 
220 volts, which was transferred by an induc- 
tion coil and then rectified. 

This higher potential made it possible to 
raise the conducting network to 16 ft. from 
the ground, thus permitting of easy cultiva- 



tion without lessening the beneficial effect of 
the current. 

Preliminary experiments upon wheat at 
Gloucester having been very favorable. New- 
man subjected 11 acres to treatment. The 
overhead network consisted of stout tele- 
graph wires mounted upon poles in rows 102 
yards apart, the distance between successive 
poles being 71 yards, and thin galvanized 
wires stretched 12 yards apart crosswise to 
act as discharge wires. A difference in the 
rate of growth was noticeable vers- early, 
and at harvesting the straw averaged from 
four to eight inches taller, and the Canadian 
wheat ripened three or four days sooner. 
The yields were 39 per cent better for Cana- 
dian wheat, and 29 per cent better for English. 
Further, the electrified wheat sold for 7.5 
per cent better price on account of its supe- 
rior quality. 

Breslauer. 1 who has written a critical 
review of the subject up to 1910, and kept in 
close touch with the progress of the work in 
Germany, tells (1909) of the results obtained 
at Halle by Kuhn, and at Holstein, Neumark, 
and Westpreussen. 

At Halle experiments were made under 
various conditions of fertilization and irri- 
gation upon a total area of about 14 acres, 
besides the control areas. This field installa- 
tion was also raised to 16)4 ft. above the 
ground. The good effect upon rye was 
already noticeable in June. It was observed 
here especially that when the wind blows the 
effects of the treatment are felt from 10 to 
lti ft. and sometimes 50 ft. beyond the limits 
of the field experimented upon, and whenever 
the control fields are adjacent, reduces by so 
much the apparent improvement due to 
electrification. This wind effect was also 
noted in work at Holstein. 

After the completion of these experiments, 
a year later. 1910, Prof. Kuhn, 14 the German 
"Nestor of agriculture," under whose im- 
mediate supervision they were conducted, 
was not enthusiastic as to the results. He 
stated that little was to be expected from the 
English procedure, as the advantage apparent 
during growth did not appear in the yield. 
His control fields of grass and grain gave the 
r results. Only fodder and sugar beets 
bettered, the latter indeed having an 
increased sugar content. Clover and cab- 
bages gave uncertain results. He considered 
that would demand at least a 15 per 

cent increase in yield. 

Breslauer 1 concludes that the investiga- 
tions already made show that the process and 

apparatus is entirely practicable. He esti- 
mates the cost of an equipment for 61.8 
acres as follows: 

Generating apparatus $595.00 

Field equipment 595.00 

Power consumption, 5kw-hrs. per day 

(at 5c) = 25c, for season, 150 days = $ 37.50 

Interest on $1190.00 at 5 per cent. . $ 59.50 

Sinking fund at 7 per cent 83.30 

Repairs at 2 per cent 23.80 

Power 37.50 

Labor (one man two hours a day). . . . 47.60 

Total $251.70 

Medium to poor yield from wheat: 2000 
lb. per acre, 

For 61.8 acres $23S0.00 

30 per cent increase 714.00 

Profit $714.05— $251. 70 = $462.30. 
Ordinary profit from 61.8 acres = $71.40. 

In a later contribution Breslauer 15 describes 
the measurement of current and power con- 
sumption by typical installations at Hoppe- 

A movable coil ammeter of great sensitive- 
ness was inserted in the ground wire. The 
order of magnitude of the voltage was deter- 
mined by measuring the length of spark in 
the air, it being known that between balls 
of 25 mm. diameter it requires about 3000 
volts per mm. to produce a spark. 

In dry, and not extremely hot weather, 
with an east wind, the voltage averaging 
about 65,000 volts, he estimates that, allow- 
ing for a certain inequality of distribution, 
the current for every 10 sq. ft. is about 0.43 
X 10~ 5 milliamperes. 

Hence the energy consumption is about 
0.26. 10" 3 amp. X 65,000 volts =17 watts = 
0.28 10" 3 watts per 10 sq. ft. 

This is from 1000 to 10,000 times the 
transfer of electric energy occurring naturally 
during a year, as estimated by Kahler. 16 

Gerlach and Erlwein 11 give an account of 
agricultural experiments upon the Kaiser 
Wilhelm Institute of Agriculture Experi- 
mental Grounds at Mocheln for which the 
equipment was supplied by the firm of 
Siemens & Halske. 

The electrical treatments included high 
tension static electricity, making the net 
positive in some cases, and negative in others, 
and high tension, single-phase alternating 

The network consisted of a heavy galva- 
nized wire supported on well insulated poles 
around the outside of the field, and suspended 



from this, across the field, thin galvanized 
iron wires at a height of 20 ft. 

The electrical equipment consisted of a 
four-horse power alcohol motor belted to a 
direct-current dynamo, and a transformer. 
The two influence machines were run In- 
direct-current motors. 

The experimental plots comprised an area 
of 800 sq. yds. besides control plots of one-half 
this area located at a distance of 330 ft. The 
plots were treated with various kinds of 
fertilizer, some were irrigated and others 
not. The crops included cabbages, barley 
and oats. 

The alternating-current antennae averaged 
a voltage of about 20,000, the static antennae 
30,000 volts. The power consumption for 
the former was about 770 volt-amperes, for 
the latter about 30 watts. The irradiation 
was begun after planting, and continued 45 
days continuously day and night. No 
difference was apparent between the electri- 
fied and untreated plants, though there was 
a considerable difference between the watered 
and unwatered, and between those differently 
fertilized. Mention is made of the occurrence 
of a drought. The harvest, occurring 120 
days after sowing showed practically identical 
yields for treated and untreated plants, with 
slight evidence of injury by the alternating 

The account gives the most extreme detail 
of electrical outfit and arrangement, but is 
vague as to the weather conditions, etc., 
which other investigators have found so 

Hostermann, 17 1910, used a network of 
telephone wires from 63^2 to 8 ft. above the 
ground and 13 ft. apart, and obtained his 
current from the atmosphere by means of 
a steel cable 820 ft. long, supported by a 
balloon or by several kites. He estimated, 
having an instrument reading to only five 
volts, from other measurements, that he got 
a potential of about 25,000 volts. This 
method gave him the best results of any, 
increasing the yield on various crops from 15 
to 40 per cent. He found that the atmos- 
pheric potential gradient varied with the 
season, the time of day, the temperature, 
and the weather, reaching maxima from 
December to February, shortly after sunrise 
and just before and during dusk, at low tem- 
peratures, and during fog, snow, hail or rain 
and especially during thunderstorms. 

The conditions under which treatment is 
applied are important, it being very essential 
that there should be moisture in the air as 

irradiation during dry and sunny weather 
often results injuriously to the plants. The 
most favorable times for treatment corre- 
spond with those of maximum potential 
gradient, i.e., very early morning and evening, 
and especially during a fog. He points out 
that the climate of England is especially 
adapted, and should give good results, espe- 
cially as the treatment seems to compensate 
in part for lack of sunshine. 

Exclusion of the influence of atmospheric 
electricity reduced the yield nearly 15 per cent. 

Hostermann, also using high potential 
pulsating direct current from a dynamo 
machine and transformer found that extended 
treatment was of little, or injurious effect, but 
more moderate application increased the yield 
in some cases 25 per cent. The crops treated 
included strawberries, spinach, lettuce, rad- 
ishes, etc. 

Stahl, 18 1911, claims he was able, using 
electrical stimulation, to bring a crop of 
corn to maturity after the winter wheat was 
reaped on July 25. He used a direct-current 
potential of about 250,000 volts (600 cycles) 
stepped up from a 60-cycle, 110-volt line 
and rectified mechanically. The wires were 
mounted eight feet from the ground, and 
two to three feet apart. The treatment was 
applied to one acre morning and evening, 
and the electric bills averaged two to three 
dollars per month. A variety of vegetables 
were treated. All matured much more 
quickly and resisted drought better. Only 
qualitative results are given. 

Gloede 19 used the treatment in growing 
flowers and found greatly increased vigor as 
well as resistance to harmful fungi. In a 
small outdoor plot 20 feet square he ripened 
362 muskmelons from seed in less than nine 
weeks, and the fruit was noticeably sweeter 
than usual. 

An installation near Prague, 20 designed by 
Breslauer, operated upon an area of 89 acres 
by means of a network of iron supported 
by porcelain insulators upon wooden poles at 
intervals of 328 feet apart across which was 
stretched a network of 0.008-in. wire at a 
height of 13 feet above the ground. Direct 
current at 120 volts, 2 amp., was supplied by 
means of a mercury interrupter, a trans- 
former, producing 100,000 volts, and a rectifier. 
The network was always made positive, and 
the treatment applied only a few hours each 
day, being always discontinued in case of 
rain, which caused leakage, and of great heat, 
under which latter condition the current is 
injurious. In spite of an unusually dry 



season yields in some cases double that of the 
control plots were claimed. Details as to 
sort of crop and actual yields are not given. 

Basty. 4 experimenting on a regimental 
garden, in France claimed good results. 

Dorsey 2 applied to small greenhouse beds 
for an hour night and morning, daily, alter- 
nating current of 200,000 cycles frequency, 
at 1(1,000 volts from a Tesla machine and 
transformer, consuming about 130 watts. 
He used a network of 0.01-in. wire at a height 
of 15 in. above the bed. He found by weigh- 
ing representative plants a marked gain 
amounting to 75 per cent for lettuce. This 
method gave better results than illumination 
or earth currents. 

He next applied a silent discharge by 
means of a network of O.OS-in. copper wire, 
nine feet above the ground, 15 feet apart on 
insulators designed for 60,000 volts, to over 
an acre of garden using 10,000 to 20,000 volts 
at 30,000 cycles for five hours daily for two 
months and 50.000 volts for one month. 
Interruption of service makes the results only 
qualitative in value. Almost all of the 
irradiated plants, including radishes, lettuce, 
beets, cabbages, cucumbers, turnips, melons, 
tomatoes, and parsnips, gave a better growth 
than on the untreated acre. Beans and peas 
were affected slightly, but all the other plants 
matured at least two weeks earlier than the 
control plants. Tobacco showed a 20 per 
cent gain. 

Peaslee, 12 1913, applied 100,000 volts from 
a Wimshurst machine on wires 10 in. from 
the soil to seedlings, with results which he 
describes as disastrous, at first. Later, by 
applying the voltage only at night and on 
cloudy days he increased the growth of straw- 
berries 27 per cent, and beetroots 14 per cent, 
tops 39 per cent. He could not establish 
any optimum voltage. He found that the 
size of the wires made no difference. Cli- 
matic variations appeared to have consider- 
able effect. 

Preliminary tests with a Tesla coil gave 
qualitatively similar results. 


The impression gained from the literature 
of electro-culture is that the last word is by 
no means said. From the nature of the 
publications it would appear that the indi- 
vidual investigations have been too cursory. 
There has been too little systematic varia- 
tion of conditions, and especially of the elec- 
trical conditions. It seems highly desirable 
that a much more extensive investigation, 

providing the possibility of trying different 
intensities of electrification under various 
conditions of cultivation, irrigation, etc., all 
during the same season, should be carried out. 
It is significant that the only investigator to 
attempt an extended examination of the field 
was able to locate and eliminate many faults 
in his method, and thus obtain good results in 
the end in almost every case, often reversing 
his previous experience. If Lemstrbm, working 
with his very imperfect equipment and limited 
resources could attain so much success, greater 
development still should be possible with the 
more adaptable apparatus now available. 

The theories as to the actual mechanism of 
the action of the electric discharge upon 
plants, involve questions of physiological and 
botanical chemistry whose answers are still 
too uncertain to make their consideration 
here of profit. Lemstrom, 13 Priestly, 3 Es- 
card, 21 and Peaslee 12 discuss the subject 
briefly, and references to points more or less 
related to it are given in the bibliography 
appended to this article. 


<ii Breslauer, M. Elektrochem Z. 16, 1-5(1909) 



L'L'I S 

History. Experiments. (Method V) 
at Halle and in Holstein. Estimate 
of costs and profits of installation. 
(?) Dorsey. H. G. Elec. (L) TS. 4+2-3 (1913) 

Elektrotech Z. 35, 236-8 (1914) 

(Methods I, IV and V.) Experi- 
ments at Dayton. Ohio. 
Method IV unfavorable. Others 
i i) Priestly. J. H. Proc. Bristol (Eng.) Naturalists Soc. 

/. 190-203 (1907). 

History and account of experiments. 
(Method V) by Lodge and Newman 
at Bitton. Gloucester and Eve- 
sham. Theory and references. 
hi Charriere. G. Abstract. Elektrochem Z. 10,15(1912). 

(Methods II and V.) Note on ex- 
periments of Davidoff on Long 
Island and Bastv in France, 
(s) Rawson and LeBaron Elec. World. 47. 1067 (1906). 
Elec. (L) 57, 305-6 (19061. 
(Method III.) Favorable. 
Unconvincing reporters account. 
(«) Newman. J. E. Elec. (L) 66. 915-6 (1911). 

(Method V.) Experiments at Bitton. 
Favorable. Very brief. 
;> Cook, E. H. Elec. (L) 41. 787-8 (1898). 

(Methods II and V.) Favorable 
especially to germination. 
-i Kinney. A. S. Bull. 43. Hatch Experimental Station 

Mass.. Agricultural College. 
Detailed Abstract. Electric Engi- 
neer (N. Y.) A3. 289-92 (1897). 
(Method IV.) Laboratory experi- 
ments. Favorable. Minimum 
optimum, and maximum voltage. 
(») Kovessi La Houille Blanche ft. 223 (1912). 

Compt. Rend. 154. 289-91 (1912). 
Abstract. Elektrochem Z. 19. 224 

(Method IV.) 1100 experiments 

Compt. Rend. loo. 63-6 (1912). 
Electrochemical explanation of pre- 



(10) Schneckenberg. E. 
(n) Gerlach and Erlwein 

(12) Peaslee, W. D. 

(13) Lemstrom, S. 

( M ) Kuhn, J. 

(15) Breslauer, M. 
(ie> Kahler. K. 

( 171 Hustermann 

(13) Stahl. W. 
Gloede, R. 

(w) Cook, F. L. 

Cm) Editorial 

(21) Escard. J. 

Heber. G. 
Lodge, O. 

Elektrochem Z. 19. 151-4 (1912). 
Kovessi duplicated facts known which 
apply only to Method IV. 
ElektrochemZ. ./7.31-6, (66)-8 (1910). 
(Method IV.) Experiments at Moc- 
heln. Unfavorable. Engineering 
J. Elec. Power and Gas 88, 69-72 (1914). 
< Methods I and V.) Favorable under 
proper conditions. 
Electricity in Agriculture, 72 pp. Van 

Detailed account of experiments 
1886-1903. Favorable. Theory, 
Elektrotech Z. 31, 380 (1910). 

(Method V.) Experiments at Halle, 
(see also 1). Injurious to some 
crops, favorable to others. Brief. 
Z. Elektrochem. 16, 557-9 (1910). 
* Method V.) Energy and current 
Phys. Z. 9, 258-60 (1908). 

Measurement of electrical precipita- 
Abstract; Elektrotech Z. 31, 294-5 

(Method V.) Experiments at Dah- 
lem. Favorable under proper con- 
Elec. World 58, 1549-50 (1911). 

( Method V.) Experiments at Evans- 
ton, 111. Favorable. 
Elec. Review. West. Elect. 59, 975-6 

(Method V.) Brief account. Gloede's 
experiments. Favorable.. 
Elektrotech Z. 33, 1108-9 (1912). 
(also p. 1200). 

(Method V.) Experimentsat Prague. 
Brief. Favorable. 
Rev. gen. des. Sciences pur. et app. 
April 30. 1913. 

History and summary. Fundamental 
Electrical facts and theory. 
West Elec, 30, 59 (1902). 

( Method IV.) Small scale experi- 
ments. Favorable. 
Elec. Engr. (L) Zjg, 110-14 (1908) 

History, brief. Experiments near 
Gloucester. Favorable. 

Breslauer, M. Elektrotech Z. 29. 915-6 (1908). 

Brief review of data. 
Clark, T. Elec. Rev. West. Elec. 69, 976 (.1911). 

Improved static machine for electro- 
Guarim. E. Elec. World. 41. 554-6 (1902). 

Review of Lemstrom 's work. 
L'Eclairage Elect rique 37, 101-8(1903). 
Review of subject and theory. 
Chouchak Compt. Rend. 158, 1907 fl914). 

Effect of Method IV upon absorption 
of ammonium phosphate from solu- 
tion by live and dead seedlings. 
Bose, J. C. " Plant Response as a Means of Physio- 

logical Investigation" (Longmans 
Green & Co., 1906). 
Effect of various stimuli, including elec- 
Berthelot Compt. Rend. 131. 772-8' (1900). 

Chemical and electrical conditions 
during silent discharge. 
Lob.W. Abhand deut. Bunsen Ges. 1914. 

Abstract; Elektrotech und Maschin- 

enbau 32. 640-1 (1914). 
Effect of silent and glow discharge 
upon starch, peptone, etc., solu- 
tions. Chemical reactions. 
Elster. J. Ann. Phys. S, 425-46 (1900). 

Geitel, H. Dissipation of electricity into the 

Schneckenberg. E. Elektrochem Z. 17, 333-7 (1911). 
Elektrochem Z. IS, 5-7 (1911). 

Motion of plants and animals in the 
electric current. Review of litera- 
Waller. A. D. Proc. Roy Soc. 67, 129-37 (1900). 

Electrical effects of light upon green 
Green. R. Phil. Trans. Roy. Soc. 1SS, 188. 

Action of light on diastase. 
Pollacci. G. Atti. Inst. Bot. 2, (No. 11) 7-10 (1905). 

Influence of electricity upon carbo- 
hydrate formation. 
Bach Compt. Rend. 26, 479. 

Possible effect of electricity upon for- 
mation of formaldehyde from car- 
bon dioxide. 
Euler Ber. deut. Chera. Ges. 37, 3415 (1904). 

Finds Bach's supposition untrue in 



By A. L. Rohrer 
Electrical Superintendent, Schenectady Works, General Electric Company 

Because of the phenomenally rapid advance that has been made in the development of electrical devices 
and their applications, it is only natural that our interest is mainly focused on our present activities in the 
industry and on the future possibilities in the science. Under these conditions it is refreshing to turn and 
glance backward at the results of a series of tests made about thirty years ago to determine the relative 
merits of the pioneer engineering work of that period. — Editor. 

The celebration of the thirtieth anniversary 
of the 1SS4 Electrical Exposition (held in 
Philadelphia, last June, under the auspices 
of the Franklin Institute), calls to mind some 
important electrical tests made for the 
Committee on Scientific and Educational 
Appliances of the eleventh Cincinnati Indus- 
trial Exposition, which was held in 1883. 

This committee was headed by the very 
active Chairman, Professor Wm. L. Dudley, 
the other members being Messrs. W. A. 
Collord, Alfred Springer, F. W. Clark, and 
Joseph H. Feenster — all successful business 
men. The Committee determined to under- 
take a series of tests of the efficiency of 
electric lighting systems and so advertised 
the situation in its circulars, which were 
widely distributed. Special premiums were 
offered for (1) the best system of arc lighting, 

(2) the best system of incandescent lighting, 

(3) the best dynamo machine for arc lighting, 
i 4 ) the best machine for incandescent lighting, 
(5) the best arc lamp, and (6) the best 
incandescent lamp. 

Professor William L. Dudley selected the 
jury, which consisted of Dr. T. C. Menden- 
hall, of Ohio State University, Chairman; 
Professors H. T. Eddy and Thomas French, 
Jr., of the University of Cincinnati, and 
Mr. Walter Laidlaw, Mechanical Engineer, 
with Lane & Bodley Company, of Cincinnati. 
This jury, in turn, selected as its assistant 
Mr. A. L. Rohrer, who was then a student 
in physics at the Ohio State University. 
The jury was instructed to make such tests 
and measurements as seemed desirable and 
possible under the circumstances, and which 
would aid in arriving at a verdict upon the 
relative merits of the different exhibits. 

In response to the proposal four systems 
of electric lighting were entered for compe- 

(1) The Thomson-Houston Electric 
Company submitted a svstem of arc 

(2) The Edison Company for Isolated 
Lighting submitted a system of incandes- 
cent lighting. 

(3) The United States Electric Lighting 
Company offered a system of arc lighting. 

(4) The United States Electric Lighting 
Company also submitted a system of 
incandescent lighting. 

The exposition opened on September 5, 
1883, closed on October 6, and the jury 
was requested to make its report of the 
awards one week before the closing date. 
Several things conspired to make these tests 
less complete in some respects than was 
thought desirable by those interested. The 
time at the disposal of the jury was short, 
however, which, combined with the fact 
that the members of the jury were all engaged 
in professional work and were therefore 
unable to devote their entire time to the 
work, had much to do with governing the 
completeness of the tests. The general plan 
adopted was to have all the energy that was 
consumed by the dynamo measured by means 
of dynamometers, and all the electrical 
energy in the circuit was to be determined 
by well-known methods. The energy con- 
sumed by the lamps was also to be measured, 
as was the illuminating power. The measure- 
ments were therefore of three kinds: — 
Dynamometrie, electric, and photometric. 

It is rather interesting to relate that the 
dynamometrie measurements were deter- 
mined by the use of the cradle dynamometer, 
which at that time was a recent invention of 
Professor C. F. Brackett, of Princeton 
University, and that was the first time the 
cradle had been built in practical form (a 
small model having been built previously by 
Professor Brackett). The form of this 
dynamometer is now well known so that even 
a general description of it is unnecessary. 

It is also interesting to relate that the 
measurements of electrical energy were made 
by the use of a pair of Sir William Thomson's 



graded instruments, the ammeter and the 
voltmeter. This set of instruments had been 
imported by the Electric Supply Company 
of New York especially for the use of the 
jury, and at that time were known as the only 
reliable instruments of their kind. 

During the tests of the arc-lighting ma- 
chines, the whole current was taken through 
the current galvanometer. With the incan- 
descent system, the total current in the cir- 
cuit sometimes was as much as 175 amperes, 
so that the jury found it necessary to make 
use of a shunt, probably the first time a 
shunt had been used on such a large scale. 
This shunt was of sufficient capacity and 
resistance that about one-fifth of the current 
was taken through the galvanometer. 

The photometric measurements presented 
the most difficult problems for the jury. 
Even at that time the expression of illuminat- 
ing power in candles was considered a matter 
of uncertainty and, as the test was intended 
to be purely a competitive one, the jury- 
decided to ignore entirely the question of 
candle-power and confine itself to a com- 
parison of the lamps under consideration. 
A 16 c-p. lamp was used as a standard and 
all measurements were made in terms of 

It is not necessary in this article to go 
into all of the interesting details with regard 
to these tests. The jury was working in 
practically a new field, and there arose many 
problems which had to be solved. As a 
result of these tests, a unanimous verdict 
was agreed upon by the jury, which recom- 
mended the following to the Committee on 
Scientific and Educational Appliances: 

(1) To the Edison Company for Iso- 
lated Lighting, a silver medal for the 
intrinsic merit and superior excellence and 
efficiency of their dynamo-electric machine 
for incandescent lighting. 

(2) To the Thomson-Houston Electric 
Company, premium of $500 for intrinsic 
merit and superior excellence in the 
following particulars, namely, highest total 
efficiency, construction of lamp, and con- 
trol of system. 

(3) To the Edison Company for Iso- 
lated Lighting, gold medal for trie intrinsic 
merit and superior excellence and high 
efficiency of their incandescent electric 

(4) To the Thomson-Houston Electric 
Company a special premium of a gold medal 
for the intrinsic merit and superior excel- 
lence of their lamp for arc lighting in the 

following particulars, namely, efficiency 
and regularity of action. 

(5) To the United States Electric 
Lighting Company, silver medal for the 
intrinsic merit and superior excellence 
of their dynamo-electric machine for arc 

(6) To the United States Electric 
Lighting Company of New York, a pre- 
mium of $300 for the intrinsic merit and 
superior excellence for the second best 
system of incandescent electric lighting. 

(7) To the United States Electric 
Lighting Company, a premium of $300 
for the intrinsic merit and superior excel- 
lence of the second best system of arc 

During the progress of the tests the jury 
listened to the briefs which were submitted 
by the representatives of the various com- 
panies which had entered their apparatus. 
It might be interesting to give the names of 
these gentlemen. The Edison Company for 
Isolated Lighting was represented by Messrs. 
John W. Howell and Luther B. Steringer. 
The Thomson-Houston Electric Company 
was represented by Professor Elihu Thomson, 
Mr. vS. A. Barton, General Manager, and 
Mr. E. F. Peck, who was in charge of the 
exhibit. The United States Electric Lighting 
Company was represented by Messrs. Curtis, 
Hine, and one or two others whose names 
have been forgotten. 

A detailed description of the tests made 
by this jury, and a summary of the results 
obtained will be found in Science, Vol. Ill, 
No. 54 (the issue of February 15, 1884). 
It is very interesting because of the detailed 
reference to the methods employed in making 
the tests and because of the fact that, after 
many years of experiments with various 
apparatus and methods, it appears that the 
methods which are now thought most suit- 
able (at least for precision), are in most cases 
those that were used in these 1883 tests. 

For example, the efficiency was determined 
by the input-output method using a direct 
means of measuring the power supply. The 
cradle dynamometer was employed which, 
in modified form, is what is used today in the 
standardizing laboratory and which gives 
very satisfactory results as far as accuracy 

During the 31 years which have passed 
between the time when these tests were made 
and today, we have passed through a long 
series of methods, based on the separate deter- 
mination of "stray power" or loss, in testing 



efficiencies. A few years ago some awoke to 
a realization that we were expending much 
more time and effort to get at the result by 
indirection than would be required to go 
straight to the answer by direct measure- 
ments. The fact was also quite strongly 
brought out at the Convention of the Ameri- 
can Institute of Electrical Engineers held in 
February, 1913, that, even when the extreme 
amount of detail which is demanded is 
carefully observed, there are certain "un- 
determined losses" which are appreciable 
and which cannot be taken care of now. 

In the various discussions covering this 
point, some strongly contended that what 
we want to know ultimately is the efficiency 
and that it would be better to go after this 
directlv as we did in some instances with 
success. Eventually, this point of view will 
no doubt prevail, although just now it is 
considered to be not advisable to say much 
about it on account of the impression that 
prevails in some quarters that such processes 
are necessarily much more expensive and 
require more time than the methods that are 
now used, namely, the separate determination 
of the losses. 

In other words, the tendency now is to go 
back to the methods of this 1883 test. 

With regard to the electrical measurements, 
the situation is quite similar. The output was 
measured in some cases by a shunted ammeter. 
and the development of this idea has produced 
the most perfect current measuring instru- 
ments which have so far been known. 

For a long time after these 1SS3 tests, 
attempts were made to produce accurate 
and convenient instruments for switchboard 
use, as well as for precision testing, in which 
the total current was passed through the 
instrument. With the exception of the 
Siemens dynamometer and the Kelvin 
balance, they have all passed into history 
and. although these instruments can still be 
classed as the best which were ever produced, 
they have not remained in use for direct- 
current measurements on account of their 

In reference to the photometric measure- 
ments, the incandescent lamp standard was 

used. This is now almost exclusively em- 
ployed as a standard. We have passed 
through a long series of standards for photo- 
metric work, most of which were unheard of 
after a few years and all of which now have 
practically been abandoned. One can recall 
a few of these, such as Methvens standard, 
the amyl-acetate standard, the pentane 
standard, the Carcel standard, and others. 
Outside of a few national physical labora- 
tories and museums none of these will be 
found at the present time. 

With reference to the tests of the gal- 
vanometers, standard cells were used. These, 
of course, were not the perfect Weston cells 
which w T e now have available, but were 
Daniell cells. Working with standard cells 
found little practical support for many years 
but now we have completely returned to this 
apparatus. Potentiometers and improved 
standard cells are the recognized methods 
found in the highest class of work. In fact, the 
desirability is now being considered of making 
terrestrial magnetic measurements with refer- 
ence to the standard cells, instead of by the 
inverse process which was formerly employed. 
The permanence and sensitivity which can 
be obtained with modern cells and gal- 
vanometer is such that a magnetic needle 
suspended in a coil of known constants 
can- be used to determine the earth's mag- 
netism by the deflection obtained with much 
more accuracy than the electromotive force 
of the cell can be determined from the coil 
and independent measurements of the earth's 
magnetic field. 

Further comments would serve to bring out 
the same main point, which is that we are 
proceeding today along the same broad 
lines that were laid down in these 1883 tests, 
and that we have tried almost everything 
else in the meantime but have returned to 
these older methods. This is not due to 
chance development, but can be explained 
by the fact that those who had this matter in 
hand at this early date had more than 
ordinary appreciation of the proper way to 
attack a problem, that is, directly from the 
front, to which method we have been com- 
pelled to return. 


By Dr. Wheeler P. Davey 

Research Laboratory, General Electric Company 

In the August, 1914, issue of the Review we published an article illustrated with "X-ray" photographs 
which gave an idea of the extraordinary ability of the Coolidge X-ray tube to successfully penetrate both 
thin and thick objects. In the following article is described one of the most recent commercial applications 
of the tube, viz., that of detecting flaws in steel castings. The service that this tube may render in the art 
of steel founding can easily be realized. — Editor. 

It has always been true that as soon as a 
new tool is perfected unsuspected applica- 
tions of that tool rapidly develop. This has 
been expecially true in the case of the Cool- 
idge X-ray tube. It is planned to publish from 
time to time results of such special applica- 
tions as may come within our experience. 
Possibly the question of observing the "pipe" 
in a steel ingot by the use of the X-ray, 
thereby being able to determine just where 
the ingot should be "cropped" may seem 
still somewhat removed, at least in so far as 
commercial applications are concerned. There 
is nc inherent impossibility in the process 
however. The case now being described is a 
long step in this direction. It is the object 
of this article to describe in detail what has 
already been done in the way of an X-ray 
examination of a certain steel casting of 
which suspicion had been aroused as to its 
homogeneity when in the machine shop. 

The original casting was two and one- 
half inches thick and weighed about a ton. 
When received at the Schenectady Works 
of the General Electric Company it had 
been machined down to approximately the 
desired shape and thickness. The amount 
still to be taken from the faces was not 
more than one-eighth inch and in some 
places was only one-sixteenth inch, but 
when this was removed it was found that 
some small imperfections had been cut into. 
These extended over an area -^bout five 
inches long and one and one-half inches 

The mechanical department at once 
chiseled away a part of the surface at this 
point, and then sent the casting to the 
Research Laboratory to determine if, by 
means of an X-ray examination, it might 
be possible to reveal still other hidden blow 
holes or imperfections. 

A Coolidge tube especially made for use on 
high voltages was set up in front of that part 
of the casting where the imperfections had 
been found. An 8 by 10-inch Seed X-ray 
plate was mounted immediately behind the 

casting and the plate was backed by a large 
sheet of lead. The distance from the source 
of X-ray to the plate was 20 inches. The 
tube was excited by an induction coil with a 
mercury-turbine interrupter. The current 
through the tube was 1.25 milli-amperes and 
the potential across the terminals of the tube 
corresponded to that sufficient to break down 
a 15-inch spark gap between needle points. 
The X-ray plate was exposed two minutes. 
At the place where the radiograph was taken, 
the finished casting was about nine-sixteenths 
of an inch thick. The radiograph obtained 
is shown in Fig. 2. The casting was then 
moved eight inches and another radiograph 
made. In this way a number of exploratory 





Fig. 1. 

Diagram of Set-up for Taking Pictures of Steel Casting. 
Drawn to one-eighth scale 

radiographs were taken through different 
points of the casting. 

All the radiographs thus taken showed 
plainly the tool marks on the surface of the 
casting. All but one showed peculiar markings 



Fig. 2. Radiograph of Steel Casting. Some of the imperfections have been chiseled out of the steel. 
The chisel marks and some remaining imperfections show plainly 

Fig. 3. Radiograph of Steel Casting showing flaw in center of casting. 
The circle shows where a piece was later punched from the casting 



which were of such shape as to strongly 
suggest that they were indeed the pictures of 
holes in the interior. In the words of the 
surgeon it was decided "to confirm the 
diagnosis by making an exploratory incision." 

Fig. 4. Photograph of Top Surface of Casting at 

place where piece was punched out. Note that 

no imperfections are visible. The U is a 

punch mark to identify top of 

piece cut out 

This has proved, then, that with the proper 
X-ray exposure blow holes or cavities may be 
disclosed in apparently solid metal of con- 
siderable thickness. A careful comparison of 
the X-ray photographs and the button photo- 

Fig. 5. Photograph of Bottom Surface of Casting 

at place where piece was cut out. Note that 

no imperfections are visible at the surface 

A circular piece, one inch in diameter, was 
punched from the casting at a point where 
one of the radiographs indicated that a blow 
hole should be found. (Location of sample 

Fig. 6. Photograph of one Edge of Button which 

was cut from the Casting (see Fig. 3) showing 

position of hole. Button was rs inch thick 

shown by circle on Fig. 3.) Figs. 4 and 5 
show that the surfaces of the casting were 
entirely free from blow holes at the point 
where the button was removed. Figs. 6 and 
7 show the ends of the hole in the button. 

graphs leads to the conclusion that very small 
air inclusions are made visible; and the fact 
that the tool marks are plainly visible on the 
X-ray plate confirms this fact. 

Fig. 7. 

Photograph of Edge of Button opposite 
to that shown in Fig. 6 

Such studies point to the desirability of 
great care in metal casting where imperfec- 
tions, ordinarily invisible, are of great danger, 
and where X-ray analysis or some other 
method is not used to check them. 



By Robert W. Adams, E. E. 

The transmission line calculator described in this article was developed to simplify the calculation of the 
voltage drop and power loss in a-c. transmission lines. The author condenses the preliminary work neces- 
sarv for a graphic solution of vector diagrams and then explains how, by means of the curve and charts 
published herewith, the succeeding steps necessary to a complete solution are arrived at. — Editor. 

Perhaps the most tedious problem which 
confronts the average electrical engineer 
is the accurate calculation of voltage drop 
and power loss in alternating-current trans- 
mission lines. This calculation is one that 
frequently has to be repeated several times 
before the most economical and efficient 
design is secured, and on this account the 
orthodox trigonometric method, while not 
in itself unduly difficult, becomes very 
laborious in its practical application. 

Accordingly, there have been proposed a 
number of "short-cut" methods, designed 
to reduce the labor of computing voltage 
drop in lines of moderate length in which 
capacity can be neglected; and one of these, 
the Mershon chart, has been very successful 
in abbreviating a portion of the process 
without departing from the strict mathe- 
matical solution of the vector diagram. This 
chart, however, in common with most of the 
other graphic methods, cannot be applied 
to a specific problem until a certain amount of 
arithmetical calculation has been performed, 
and it is this extra labor which is the most 
fruitful source of error and delay. 

With the idea of shortening this labor and 
lessening the chance of error, the author has 

Fig. 1. Diagram Representing the Magnitude and Phase 
Relation of the Transmission Generated Voltage, 
Receiver Voltage, and Line- 
Drop Voltage 

condensed into two steps the preliminary 
work necessary to the graphic solution of the 
or diagram. 

(1) The first step is the determination 
of a "Transmission Factor" from the 
formula : 


kv-a. X distance in miles 
10 X kilo volts 2 

This formula combines the load kilovolt- 
amperes, the distance that the power is 
transmitted in miles, and the pressure in 
thousands of volts between the wires at the 
receiver end of the line, in such a way that 
both decimals and large numbers are 
avoided and the fraction can frequently 
be solved by mere inspection. 

(2) The second step is the determination 
of the percentage resistance and reactance 
components of the line drop (represented 












<0 A 



i 4 























Fig. 2. Curve Sheet Showing the Relation between the 

Resistance, Reactance, and Length for No. 0000 

Wire, for 60 Cycles and 18-Inch Spacing 

by the base OD and the altitude DB of the 
line-drop triangle in Fig. 1 ) . This is accom- 
plished by multiplying by K: (a) the 
resistance in ohms per mile of a single wire 
of the size selected, (b) the reactance in 



ohms per mile of this wire at the given 
frequency and spacing of conductors. 

When the per cent resistance and 
reactance drops have been determined in 
this manner, they can be applied to the 
Mershon chart or used in the trigonometric 

. BY n w AOAMS POOVfftNCE i 

Fig. 3. 

A Chart from which the Transmission Factor K may 
be obtained for Various Wires at 18-Inch 
Spacing and 60 Cycles 

solution of the line-drop diagram (it being 
noted that they refer to three-phase work 
and are to be multiplied by two if the 
circuit is single-phase). 
In order to express graphically the whole 
simplified method just outlined, the author 
has constructed a calculating device that 

As the computing scale requires no special 
explanation, we may proceed to describe the 
evolution of the wire diagram. 

Considering first a single No. 0000, B.&S. 
copper wire at 60 cycles and 18-inch spacing, 
we find that it has a definite resistance and 
reactance for a given length in miles, as shown 
in Fig. 2. For instance, a ten-mile length has a 
resistance of 2.6 ohms and a reactance of 
.5.6 ohms, as indicated by the heavy triangle. 

This triangle corresponds in shape to the 
line-drop triangle of the vector diagram, and 
can be converted into such a triangle by 
multiplying the three sides by a factor which 
takes proper account of the nature of the 

The three sides then become : 

Ohms resistance X 

V3X three-phase current X 100 

receiver voltage 
= per cent resistance drop. 
Ohms reactance X 

V&X three-phase current X 100 

receiver voltage 

= per cent reactance drop. 

Miles X 

V 3 X three-phase current X 1 0( ) 

receiver voltage 

= K. 

Fig. 4. A Transparent Chart, which when Superimposed on 

the Chart of Fig. 3 with due Respect to Power-Factor, 

Indicates the Line-Drop in Per Cent 

of Receiver Voltage 

consists of a circular slide-rule scale for 
computing the value of K, together with a 
wire diagram for locating the apex of the 
line-drop triangle, and a transparent chart to 
indicate the actual drop in per cent of the 
receiver voltage. 


Fig. 5. An Illustration of a Setting of the Chart of Fig. 4 
Superimposed on that of Fig. 3 

It follows that for any given value of A" 
(as previously determined for the given load 
by means of the computing scale), there is a 
corresponding point on the sloping line for 
No. 0000 wire which locates the apex of the 
per cent line-drop triangle representing the 



effect of this load on a circuit of this wire. 
Similar sloping lines can be drawn for the 
other sizes of wire and the corresponding 
points can be connected up to form curves, 

Fig. 6. A Photograph of a Page of the Calculator showing 

the Curves which were illustrated in part 

in Figs. 3, 4 and 5 

by means of which the apex of the line-drop 
triangle can readily be determined for any 
value of K. 

The result is shown in Fig. 3, which is a 
reduced view of one quadrant of the station- 
ary diagram of the Transmission Line 
Calculator. The resistance and reactance 
scales are omitted from the final diagram, 
as they are not essential to the practical 
working of the device. 

We can now, by means of K, locate on the 
diagram the apex of any line-drop triangle, 
and it remains to make use of this graphically 
with due reference to the power-factor of the 
load so as to indicate the true percentage 
drop in the transmission line. In other words, 
referring again to Fig. 1, we have the point B 
and wish to know the distance OC, which is 
determined in the figure by the arc BC. 

drawn with its center at .4 and intersecting 
AO prolonged. 

In order to reproduce this arc graphically 
we construct a transparent chart consisting 
of a series of percentage arcs drawn from a 
common center A, and pivot this chart at the 
lower left-hand corner of the wire diagram. 

This chart, a portion of which is shown 
in Fig. 4, is furnished with a central reference 
line which corresponds to OC and which can be 
set at any angle to the base line of the diagram 
by means of a suitable power-factor scale. 

When the transparent chart has been set 
in this manner to correspond with the load 
power-factor, as shown in Fig. 5, and the 
apex of the line-drop triangle has been 
located for the given circuit conditions as 
previously explained, we have only to find 
the arc nearest this apex and follow this arc 
to the graduated reference line of the chart. 
There the true line drop, corresponding to 
the distance OC, can be read directly in 
per cent of the receiver (or load) voltage. 

If, instead of following the arc as outlined 
above, we follow the nearest vertical line of 
the diagram to the reference line of the chart, 
we can there read the power loss in the circuit, 
expressed in per cent of the delivered power. 
This appears in Fig. 1 as the distance OL, 
which obviously bears the same ratio to the 
line AO, representing 100 per cent, as the 
resistance drop OD bears to the power com- 
ponent AE of the load voltage. 

We can, therefore, by means of this wire 
diagram and transparent chart, determine the 
line drop and power loss correctly for any 
circuit of 18-inch spacing at 60 cycles; and, 
by constructing similar diagrams for other 
common spacings and frequencies, we can 
expand the range of the device so as to include 
the whole field of transmission and distribu- 
tion at moderate voltages. 

The result appears in Fig. 0, which shows 
a complete page of the Transmission Line 
Calculator as arranged for four standard 
spacings at 60 cycles. It is equipped with a 
transparent chart which is in the form of a 
disk having at its edge a circular slide-rule 
for computing the value of K for any given 
load, voltage, and distance of transmission. 


By John D. Ball 

Consulting Engineering Department, General Electric Company 

The study of magnetics is one which in the past has received a large amount of attention. In recent 
years, however, its necessity for successful electrical design has become more and more realized; and, as a 
result, we are rapidly gaining additional information from day to day which is making itself apparent in the 
manufacture and characteristics of the machines now placed on the market. The following group of three 
short articles contains ideas that will similarly prove to lie a valuable addition to our knowledge of magnetiza- 
tion curves. — Editor. 


For designing machines and for various 
other purposes, it is often necessary to know 
the values of magnetization for higher 
induction than is usually given by test which, 
with the ordinary apparatus employed for 
testing magnetic material, is usually limited 
to an upper range of values of magnetization 
forces of from 200 to 400 gilberts. Obtaining 
higher values therefore involves special testing 
apparatus (attended with considerable cost 
and care in making the determinations) or the 
extrapolation of the curves obtained in the 
ordinary way. 

Because of the nature of the magnetization 
curve, direct extrapolation is difficult and, 
for a given induction, the value of the mag- 
netization taken from such an 
extrapolated curve is likely to 
be in considerable error. A 
rational extension to the curve 
may be made, however, by means 
of the equation that is found 
by the use of the reluctivity 
curve, which is the reciprocal of 
the permeability plotted against 
H. Such a curve approximates 
a straight line over a wide 
range of values as has been 
described elsewhere. * 

The reluctivity, p, has been 
expressed by the equation 
p — a+oH , wherein a is a 
constant representing the dis- 
tance from the A-axis to the 
intercept of the p-H curve if Fig. i. 

continued along the straight line 
and a a constant representing 
the slope of the line. Therefore, the 

This is approximately true within the range 
of ordinary test when the total of the mag- 
netic material and of the air are taken to- 
together and also when the metallic density 
/3o which is equal to /3 — H is considered. The 
true or metallic reluctivitv 

Po= a +a H 


wherein a and c are constants, calculated 
on the basis of metallic density, and in 
consequence differ slightly from a and a. 

Fig. 1 gives magnetization, permeability, 
and reluctivity curves for sheet steel plotted 
from the data given in Table I. 

It will be noted that /3 and O are practi- 
cally identical up to H = 10 and differ but 
slightly at H = 200. Taking values of H 
and p from H = 50 to H = 200 as representing 





a+a H 


» Trans. A.I.E.E., Vol. VIII, p. 485 et. seq. 
" Engineering Mathematics." Steinmetz. Ed. 1911, p. 294. 
General Electric Review, Vol. XVI, 1913. p. 750. 

Curves of Magnetization, Permeability, and Reluctivity for sheet steel 
plotted from the data furnished in Table I 

the straight line, we obtain by the 2 A 
methodf the equations : 

p = 0.00058+0.0000475 H 

S= * 

p 0.00058 + 0.0000475 H 

+ "Engineering Mathematics," Steinmetz, Chapter VI. 



p„ = 0.00062 +0.0000479 H 

a= H 

Po 0.00062+0.0000479 H 

These equations show close agreement 
between p and p . 

Theoretical considerations prove that the 
use of po is much better, as the curve 
representing po — H may persist as a straight 
line; whereas the curve p — H must con- 
tinually bend downwards and have the 
curve p = 1 as its asymptote (otherwise it 
would show the material to become diamag- 
netic at very high inductions and eventually 
possess infinite reluctance or zero permea- 
bility, which cases could only be true if 
metallic densities are considered). 

When obtaining extrapolations of the fi — H 
curve, by the means outlined above, (as 
£ and not /3 is desired in the curves used for 
design purposes), it might be a temptation 
to use equation (1), but if such is done, we 
would become involved in considerable incon- 

sistencies. Therefore, when extensions of the 
ji — H curve are desired, it is necessary to use 
the equations of 0o and to add to the results 
obtained the values of H, which gives the 
equation : 

<>-db +fl (3) 

A study of the equations will readily show 
the advisability of using the latter equation. 
When high values are reached a or ao 
becomes negligible and the equation becomes 

0= —, which is to the effect that a final value 

of saturation is reached which, in amount, is 
the reciprocal of the slope of the p — H curve. 
As a matter of fact /3 does not represent an 
absolute saturation value as, after the 
material is saturated, /3 continues to increase 
by the same amount as the increase in H. 

Fig. 2 gives curves showing the plotted 
results of Table II. These data were calcu- 
lated by equations (1) and (3). 













































































Results for p 
obtained by 
equation (3) 

Results for /3 
obtained bv 
equation (1) 


Per Cent 


















• 5.9 



























It will be noted that the per cent error in 
at H= 10,000 is not large, but that if the 
errors of H at a given value of /3 were con- 
sidered, the errors become enormous. 


In extrapolations of mag- 
netization curves the equation 

introduces large er- 

enon, but is largely a function of the scale 
selected to represent the magnetizing forces. 
This fact may be illustrated by reference to 
Fig. 3. The data from which these curves 

necessary to use 
-H or some other 

^ a+aH 

TOTS. It is 

a +<T H 
form that provides for the metallic 
density to which the H values are 


When discussing magnetization 
curves we frequently hear state- 
ments that machines are designed 
at inductions which are defined 
with reference to the "knee" of 
the saturation or magnetization 
curve. As example : In such and 
such an apparatus the flux density should 
be below the "knee," in a certain magnet the 
design is to have the flux density well on or 
above the "knee," or the "knee" of the curve 
for a certain magnetic material occurs at a 
certain densitv. 

800 900 

160 ISO 

40 45 

































Fig. 3. Successive Portions, plotted to differing scales, of a single Magneti- 
zation Curve. Note the startling fact that the "knee" appears 
to be located at different and [I values 

were plotted will be found in Tables I and II 
of Section I of the present article. A glance 
at Fig. 3 will readily show that the point of 
maximum curvature (which is usually the 
interpretation of the "knee"), occurs at 
different densities on the different curves, 
which are all plotted from the same 
data, and differ only in that different 
scales for H are employed. 

The "knee," if derived from these 
curves, would be in the vicinities of 
the following values: 

Curve No. 

Flux Density at "Knee" 




Fig. 2. Extrapolations of Magnetization Curve plotted from 
the calculated results of Table II 

This conception of matters has beyond a 
doubt led into many errors and uneconomical 
designs due to the fact that this so-called 
"knee" is not entirely a magnetic phenom- 

Fig. 4 shows the second curve of 
Fig. 3 taken alone in which the 
above induction values are empha- 
sized by small circles. Fig. 4 shows 
there are no evidences in this case of 
bends which are shown by plotting to the 
other three scales. Referring to Fig. 5 and 
Fig. 6, wherein the data are plotted on 
logarithmic and semi-logarithmic paper re- 



spectively, we have curves in 
which equal percentage of in- 
crease of H gives equal abscissae. 
Here we have no evidences of 
"knees." at points in the above 
tabulation, which further proves 
the point. The top curve of 
Fig. 3 is steeper at the high 
inductions than the other three. 
due to the fact that the iron is 
saturated and the increase of /3 
is mostly due to the air. 

The above discussion applies 
to related curves, as for example, 
wherein line voltage of machines 
is plotted against field current. 

The conclusion is that the so- 
called "knee" of the curve is a 
mechanical bend, the position 
of which is due largely to 
the scale selected. This fact —P~ 
should be carefully consid- 
ered when interpreting 
curves of this nature. 








_ L 




— = 

























Fig. 4. The Complete Magnetization Curve shown in parts in Fig. 3. 
circles indicate the locations at which the "knee" occurs if curve 
was drawn to each of the scales used in Fig. 3 



Owing to the shape of 
the regular /3 — H curves 
used for design purposes, 
the curves are usually divi- 
ded and plotted to several 
scales to facilitate clear- 
ness. If this is not done it 
is difficult to determine 
values at low values of H 
up to the mechanical bend. 
or so-called "knee," unless 
results are calculated from 
a permeability curve. It is 
likewise difficult to deter- 
mine values of H at high 
values of /3. Several scales 
are likely to lead to con- 
fusion and the number of 
changes of scale is thus 
limited. If only one scale 
be used, we have a typical 
curve such as shown in 
Fig. 4. This figure is drawn 
from data in Tables I and 
II in Section I of this 

It is desirable to select 
a method of drawing mag- 
netization curves which 




— r- 





— - 

— L 









30 40 50 60 70 SO 100 


loo xo too eoo too 

Fig. 5. A Magnetization Curve plotted on paper having logarithmic abscissa 
and ordinate scales 'same values of tf and H used as in Fig. 4) 


— i — r— T 

1 1 









1 1 

^L^— ■ 


i : 

' 1 

I — \— 

+ ~ 




r ~^r* 


' 1 1 


^ — 

*" ! 

! | 


1 I 








! ' T" 

" T- 

""LT " 



f i 


u -r 

i — f— 







i i 



* S 6 

78 tO 




40 X60 

70$o no 


10 A 





Fig. 6. A Magnetization Curve plotted on paper having a logarithmic abscissa 

scale and a uniform ordinate scale (same values of 

and // used as in Figs. 4 and 5^ 


will give a wide scale for H at low values, 
and which will cause the scale to be con- 
tracted at high values. On general prin- 
ciples, it is also desirable to obtain a curve 
having a general slope of approximately 
45 deg. 

A method of obtaining such a curve as 
above outlined would be to plot the data on 
logarithmic paper. Such paper is on the 
market and is easily obtained. On such paper 
the vertical and horizontal scales are both 
divided according to logarithmic progression. 
Plotting the data of Fig. 4 on such paper gives 
the curve shown in Fig. 5. This draws out 
the H scale in an excellent manner but the 
arrangement is poor considering the ordinates, 
as at high values of /3 the scale is contracted 
and at low values it is needlessly exaggerated. 
A better scheme for plotting such a curve is 
to use paper with a logarithmic scale for the 
abscissae and a straight, or even-division 
scale, for the ordinates. The data of Fig. 4 

so plotted are shown in Fig. 6. This gives a 
very desirable H scale and a very good scale 
for 0. It also gives a good mechanical slope. 
Paper of this ruling possesses a great advan- 
tage in that any desirable scale may be used 
for ordinates, whereas for a logarithmic 
scale in both directions, the choice is much 
restricted. Considering Fig. 5 we have a 
single very readable curve. The knee of the 
saturation curve is not so pronounced in this 
case as when plotted on regular even-division 
paper, but this constitutes a further change 
in favor of the logarithmic paper because, as 
has been shown in Section II of the present 
article, the knee is not in reality a definite 
point as its location depends upon the scale 
employed when plotting the curve. Fig. (i 
contains three blocks of abscissae. An addi- 
tional block either way would give either a 
very wide H scale at low values or a very- 
contracted one at high values, which is in 
accordance with what is desirable. 


By Theodore Bodde 

The advantage accruing from the employment of specially trained men is fully realized by most large 
manufacturers, and the results in improved products and more efficient service from such employees have led 
many concerns to introduce the so-called apprenticeship course, which is simply a school for the education 
of the workmen in the fundamentals, theoretical and practical, on which a particular industry is built. The 
essentials of such a course are outlined in this article. — Editor. 

In a large factory building belonging to the 
General Electric Company, at West Lynn, 
Massachusetts, there exists a unique school of 
practical electrical and general technical 
knowledge; unique, because it combines and 
mingles intimately the practical factory at- 
mosphere with the theoretical ether of science. 

This educational institution, commonly 
called an apprentice system, gives practical 
instruction through factory work and theo- 
retical knowledge through class room lectures. 
The class room work is so arranged as to 
occupy slightly less time in years than the 
practical work. If a student therefore fails to 
pass in one of the classes and is obliged to 
repeat it, he can still finish all classroom 
work within the prescribed time limit of 

The educational institution provides a c - 
well for young men with no more than a 
grammar school education as for high school 
graduates. The grammar school graduates 
are placed in the so-called apprentice school, 
while the high school graduates enter the 
engineering school. They are selected out of a 
large number of applicants. 

In the apprentice school, the young men are 
developed into efficient skilled workmen, 

assistant foremen and foremen, and tool 
designers. In the engineering school they are 
converted into efficient practical engineers. 

The classroom work of the apprentice 
school stretches out over a period of nine terms 
of 14 weeks each. That of the engineering 
school covers a period of seven terms of 14 
weeks each. There are three terms in each 
school year. 

The teaching staff consists of six instructors 
and one superintendent. 

The above outlines in a general way the 
system of this technical school. We will now 
consider some of the different principles and 
methods which are followed in the electrical 
department, in order to give a general idea 
of the system. 

To the apprentices one term of electricity is 
given, while to the students of the engineering 
school five terms of electricity are allowed. 

The first principle followed, for the purpose 
of effectually impressing the mind of the 
young man, is that of concrete representation 
of the different truths which are taught him. 
The reason for this lies in the fact that these 
young men have generally left school at an 
early age. Consequently, theory and its 
demands on the imagination are almost 



unknown to them and the imagination has 
not been trained to its full strength. On the 
other hand, having been in contact with 
matter and material things during the greater 
part of their lives, they can, with no difficulty 
whatever, see through material things and 
material representations, where they would 
be powerless were those representations only 
abstract ones. Therefore, it is through this 
concrete method of representing things that 
one must appeal to them. 

Technical education consists in impressing 
on the mind the relations between natural 
phenomena; in other words, in leading the 
pupil to discover the links connecting different 
facts. In popular language this discovery is 
expressed by the saying: "I see," which 
means nothing else than "I see the link; I 
understand;" for when we see a thing we 
understand it. Now if this connecting link 
can be made visible by means of really visible 
things, instead of by things which are only 
visible through an effort of the imagination, 
we shall be able to make all things under- 
standable to those whose imagination is not 
strong enough for that effort, and technical 
education for the masses will become possible. 

It is true that education consists also in 
training that very effort of the imagination 
which is needed for the concentration of the 
mind. It is this branch of education which 
produces thinkers. It produces, however, 
perfect fruit only when applied to the very 
few who have a natural aptitude for thinking. 
The large mass take up only facts and relations 
and become effective tools, but very few 
among them become thinkers and leaders. 
In our present civilization it is well that this 
should be so. At the same time we may long 
for some future in the advancing ages when 
this condition will no longer be necessarv, 
and everybody will be trained for the beautv 
and development of himself and of the race. 

At present, the world needs many tools for 
its material growth, and the General Electric 
Company, which is itself a small world, 
daily feels the need for efficient tools, and 
all efforts are exerted in this direction. If, 
now and then, thinkers are mixed among the 
tools, they will be recognized sooner or later, 
and will step out of the mass through their own 
efforts. Hence, for the present, the methods 
of education should not be molded for them 
but for the large masses. Neither should the 
methods of education be molded for the other 
extreme, the dummy. In the General Electric 
apprentice and engineering school, the dum- 
are eliminated while passing from the 

lower classes to the higher ones. This is 
done by a simple weeding out process, 
through keeping a close observation and a 
just record of their doings and progress 
throughout each term. An educational com- 
mittee meets every week and carefully elimi- 
nates the chaff from the wheat, the result being 
that the higher classes are very nearly perfect. 

The following are some examples of the con- 
crete representation of things as applied in the 
electrical engineering department of this 
school : 

Throughout the first term, the text book of 
W. H. Timbie. "Essentials of Electricity," is 
used. The beginners have special trouble in 
grasping the idea of line drop in transmission 
lines and other similar very real and impor- 
tant phenomena in power transmission. The 
reason is obvious. Transmission lines are so 
large that they have never been grasped in 
their entirety in the imagination of the 
student ; they are too long to be contained 
in the narrow space of his vision. In order to 
overcome this difficulty, a miniature trans- 
mission line was made, reproducing in every 
way the phenomena of a large power trans- 
mission. The lines are made of thin resistance 
wire and are stretched across the whole length 
of two blackboards which run along the wall of 
the classroom. A set of incandescent lamps at 
about the middle of the line produces one 
load, and another set of lamps at the end of the 
line produces another load there. The begin- 
ning of the wires are switched to two binding 
posts, between which are 275 volts d-c. 
By varying the number of lamps, different 
loads are put on the line at different points, 
thus producing different currents and different 
line drops in the sections of the line. These 
values are measured in a direct way by the 
students. This gives them practice in the 
manipulation of d-c. voltmeters and amme- 
ters. The readings are then written down in 
chalk directly over the corresponding sections 
on the blackboard. From these results, 
calculations are made relating to power loss in 
the different sections of the line, voltage on 
the loads, power delivered to the loads, total 
power transmitted, etc. All these calculations, 
written dow T n again on the blackboard over 
the corresponding part, are then finally 
checked up by means of direct measurements. 

The three-wire balanced and unbalanced 
systems are also reproduced in miniature by 
the same means, and the general run — first 
measurements, then calculations, and at last 
checking up by other measurements — is 
essentiallv the same as before. 


It is remarkable what good results this 
method of teaching has produced. 

In a latter part of the term the d-c. gen- 
erator and motor phenomena are illustrated 
by means of an old fashioned bipolar shunt 
wound dynamo which has been fixed for the 
purpose and provided with a flywheel. This 
makes possible the illustration of the counter- 
electromotive force which exists in a running 
motor. Suppose the dynamo has been 
connected up at the end of the above de- 
scribed miniature transmission line, and runs 
as a motor. A set of lamps on the same trans- 
mission line shows its bright lights as a 
result of the power which it takes from the 
same source. If now the double-pole switch 
between the binding post and the trans- 
mission lines is suddenly opened, the motor, 
because of the inertia of its flywheel, will 
become a generator, and the lamps will 
still show their bright lights, this time, 
however, taking their power from the dynamo 
side; for the current on that side is reversed, 
as is clearly shown by means of an ammeter. 
The voltage on the line can be measured at 
the instant that the double-pole switch is 
opened, which serves to illustrate in a clear 
and real way the counter-electromotive force 
which existed an instant before while the 
dynamo was still running as a motor. 

Thus the student becomes familiar with all 
the secrets of the dynamo. Even this 
counter-electromotive force, so often the 
stumbling block to beginners, becomes visible, 
almost palpable to them, and impresses 
itself on their minds. The measurements of 
voltage and currents, in relation to the speeds 
through which the flywheel passes, are then 
written down, calculations are made and 
again experimentally verified, and it is thus 
that the different phenomena enter into the 
mind almost without effort; for the student is 
interested in these different operations from 
start to finish, and is not tired out by an 
undue effort of the imagination. The chan- 
nels between his senses and his mind are wide 
open, and the knowledge enters without effort. 

During the second term of electricity, 
Swoope's textbook, "Lessons in Practical 
Electricity," is used. This textbook is rich 
in material, and in this lies its great merit, 
for it offers many topics to be treated and 
talked about in the classroom. It describes 
many experiments, and to follow these de- 
scriptions requires a certain amount of the 
student's imagination. It is to be noticed 
again that the student of the engineering 
school is a high school graduate and has had 

his imagination trained originally to a greater 
extent than has the average apprentice. As 
the time is limited, considering the large scope 
of the book, this term is mainly devoted to 
theory, though here and there concrete 
illustrations are made if the described experi- 
ments of the book do not convey the fact 
clearly enough to the mind. 

The third term of electricity is devoted 
entirely to experiments and laboratory work. 
Large d-c. and a-c. dynamos and the necessary 
instruments are put into the student's hands, 
and under the direction of their instructor 
they make the usual practical tests relating to 
voltage, speed, load, losses and efficiency. It 
is surprising how quickly the students get 
hold of this term's work and of the right way 
of doing things. Their enthusiasm and 
pleasure in the work is very visible in the 
neatness with which they make up their 
reports. Some of these are almost pieces of 
art, so carefully are the sketches drawn and 
the curves traced. 

After this term of heavy practical work, 
the student goes back again to pure theory. 
Two terms of advanced electricity along the 
lines of Franklin & Esty's textbook of 
electrical engineering now follow. During 
this time the student has ample occasion to 
verify and think theoretically over the 
different points and phenomena which have 
come up during the former term, and thus the 
last foundation stone of electrical knowledge 
is deposited in his brain. 

The classroom in which the student gets 
these advanced courses of electrical engi- 
neering is in the laboratory, so that the 
whole atmosphere is impregnated with the 
practical developments of the great industry. 
Dynamos, rheostats, and all kinds of motors 
look at him from all sides while he ponders 
over some intricate problem, and like real 
friends suggest ideas to him. The walls carry 
charts illustrating such useful rules as the 
famous Fleming's three-finger rules for motor 
and generator directions, and the unconscious 
daily look at these charts produces on the 
student's mind a lasting impression, in the 
same way as in daily life the advertising 
poster impresses the public mind. If there 
is any formula, any figure, difficult yet 
useful to remember, there is no better and 
easier way of mastering it in one's memory 
than by posting it in some conspicuous 
place to which the eye is turned every day. 
These repeated impressions will leave their 
mark without requiring any acrobatic effort 
of the brain. 




By W. C. White 

Research Laboratory, General Electric Company 

Radio communication, the term recently adopted by the profession in lieu of the popular but inappropriate 
word "wireless" can be divided into two general classes — telegraphy and telephony. During the past ten years 
or more many experiments in radiotelephony have been recorded, but communication by this means has never 
enjoyed the practical or commercial applications that are common to radiotelegraphy. This article describes 
the principles involved in radiotelephony, and the difficulties encountered which have so far kept it more or 
less in the experimental stage. — Editor. 


Shortly after radiotelegraphy had become 
an accomplished fact, radiotelephony was 
proposed and experiments undertaken along 
that line. The fact that apparatus was at 
hand by which the voice could be made to 
give a variable form of electric current and 
apparatus by which this form of current could 
be made to reproduce the original voice made 
the problem seem simple in comparison with 
the original development of the telephone. 

In order to understand the problems of 
radiotelephony and why it has made so 
much slower progress than radiotelegraphy, 
it is necessary to review some of the principles 
involved in radiotransmission in general. 

Fundamentally, what happens in radio- 
transmission is that a certain amount of 
energy is generated and liberated at the 
transmitter, from which it radiates more or 
less in every direction, and a very minute 
portion of it is intercepted at the receiving 
station where it energizes the receiving 
apparatus. This is therefore a transfer of 
energy and as such requires a medium. 

It has been shown that light and radio- 
waves are similar phenomena in this medium 
which we call the ether. Now as to how these 
radio or electromagnetic-waves are set up in 
the ether, a rough analogy will help to make 
the theory clear. Imagine a paddle dipped 
vertically into a body of water; if this is 
moved very slowly back and forth water will 
merely flow from the volume in front of the 
advancing paddle around the edges to the 
space just vacated. Floating corks arranged 
in a circle about the paddle and at a radius 
of several feet away would not show any 
movement, proving that all the energy used 
in moving the paddle was expended as fric- 
tion at its surface or in eddies or currents set 
up in the immediate vicinity. 

Now suppose the frequency of the back 
and forth motion is increased. Common 
experience tolls us that waves will be set up 
which hi ppreciable after a certain 

frequency of motion has been passed. Bodies 

floating in the water at a distance will be 
moved up and down (even against an applied 
friction) as the waves pass them, showing 
that in the case of the rapidly swinging 
paddle its energy is used up in two ways; 
in the first place, as friction, as already 
mentioned and, in the second, by waves 
which transfer the energy through the medium 
away from the source until it sets some mass 
swinging whose friction dissipates the energy 
transmitted. Naturally, if one wishes to 
make short length waves a small paddle would 
be moved rapidly, and for long wave lengths 
a large paddle moved slowly. 

In order for a medium to transfer energy 
away from a source by wave motion, it must 
have inertia; and of the tangible mediums 
with which we are familiar, the less their 
density the higher the rate of vibration must 
be before an appreciable portion of energy 
leaves the source by means of wave-motion 

Returning now to electromagnetic-waves: 
If a straight conductor in space is carrying 
an alternating current at GO cycles frequency, 
the only loss we could measure would be that 
due to its resistance. It is trite, of course, 
that if a conductor of a second closed circuit 
were to parallel the first conductor, a current 
would be induced in the former which would 
consume energy. This is due to a magnetic 
field about the first conductor which may be 
said to grow from and collapse upon its 
source twice each cycle, but never traveling 
away from it continuously. 

If a conductor carries a current at say 
100,000 cycles, energy in the form of electro- 
magnetic-waves will leave it and travel away 
with the velocity of light, and, if these waves 
intercept another circuit, a current of 100,000 
cycles frequency will there be set up. The 
amplitude of these waves- decreases as the 
distance from the source increases; and ex- 
perience shows that a certain loss of energy 
occurs as the waves travel in space, due, 
undoubtedly, to atmospheric conditions, which 
loss is termed "absorption." 



It is not to be inferred from these argu- 
ments that there is any theoretical reason 
why water waves or electromagnetic-waves 
of very low frequency cannot be produced. 
For instance, a huge barrier 1000 miles long 
moved through an amplitude of, say 1000 
miles, in the middle of the Pacific ocean with 
a swing once a day, would set up waves of 
enormous power, probably causing a tidal 
flood on the shores of the ocean. 

In a similar way, a huge electrical capacity 
in the transmitting radiating circuit, charged 
at an enormously high potential, would 
radiate waves at a frequency of 60 cycles. 
It is impractical, however, to construct an 
aerial radiating system of sufficient capacity, 
and corona losses prevent the utilization of a 
high enough voltage. 

The simplest and most commonly employed 
method of obtaining high-frequency currents 
is by spark excitation. A weight suspended 
by a spring will have a natural period of 
vibration, depending upon the stiffness of the 
spring and the mass of the attached body. 
It will take up this frequency of vibration 
when struck an upward or downward blow 
and continue its oscillations for some time. 
In an analogous way an electrical circuit 
having inductance and capacity will have a 
high-frequency electric current set up in it 
when its circuit is completed by a spark 
which allows readjustment of the charge 
stored in it. 

As mentioned, a high-frequency current is 
set up in the circuit of the distant receiver, 
due to the aerial wires (these intercepting 
the electromagnetic-waves from the trans- 
mitter). As these currents are minute, the 
most sensitive form of current indicator must 
be employed. 

In order to get an idea of the magnitude 
of the currents and the amount of energy 
involved, a few quantitative examples will 
be given. The radiating circuit of a radio- 
transmitter is said to have a certain number 
of ohms resistance which may be defined as 
that quantity which when multiplied by the 
square of the current in amperes gives as a 
product the number of watts dissipated. A 
large station designed to transmit a distance 
of 1000 miles or more may use 75 amperes 
in a circuit of 8 ohms so-called antenna 

In a receiving station 2000 miles away, 
having a resistance of 25 ohms in its receiving 
circuit and apparatus, a current as high as 
•50 microamperes may be set up, which means 
about 6X10 _S watts." When we consider the 

distances, this current seems large; the 
pointer of the usual type of sensitive, portable, 
direct-current voltmeter will give about a 
one millimeter movement with such a value 
of direct current. It is to be remembered, 
however, that in the receiver circuit the 
capacity and inductance are adjusted for 
resonance for the incoming frequency, so that 
in order to realize this amount of current in 
any indicating device, its resistance must be 
very low. 

Now for direct currents the d'Arsonval 
galvanometer principle, such as employed in 
most direct-current indicating instruments, 
is the most practical form of sensitive current 
indicator. For alternating currents of fre- 
quencies of about 150 to 2000 cycles, the 
Bell telephone receiver is most sensitive and 
simple. A good telephone receiver is respon- 
sive to one-tenth of one microampere alter- 
nating current at a frequency of 500 cycles. 

The currents induced in the receiving cir- 
cuit, from a transmitter generating its oscilla- 
tions by spark discharges as mentioned, con- 
sist of groups of very high frequency current 
coming at intervals determined by the rate 
of spark discharge. 

It is evident, then, that in the receiving 
circuit some device must be utilized which 
will respond to the minute high-frequency 
currents set up, or they must be transformed 
into a form of current to which a galva- 
nometer or telephone is adapted. 

This latter method is most commonly 
employed, several rectifying devices being 
available which change the high-frequency 
groups more or less perfectly into a half -wave 
alternating current having the same frequency 
as the spark intervals at the transmitter. 
Such a form of current will actuate a galva- 
nometer or give response in a telephone, the 
latter being ordinarily used because of greater 
convenience and speed of operation. In 
practice, the sparks at the transmitter are 
made to occur at rapid and regular intervals, 
so that a fairly pure musical note is heard in 
the receiving telephone. 

Differenl spark frequencies will produce 
corresponding tones in the receiving 'tele- 
phones. This really illustrates the funda- 
al principle of a radiotelephone trans- 
mitter, viz.. the radiating of high-frequency 
electromagnetic-waves in groups correspond- 
ing to the tone to be transmitted. 

Modifications Necessary for Telephony 

In order that the voice may be reproduced 
in an ordinary Bell telephone receiver, a 



current must be passed through its winding 
which has an alternating-current component 
corresponding in frequency and wave shape 
to the fundamental and overtones in the 
voice, and in amplitude to its loudness. 

Thousandths ofa Second 

Fig. 1. A Series of Direct-Current Waves which are made 

up of a 1000-Cycle Alternating Current 

and a Continuous Current 


40 SO 

Mi/Z/oM/ts of a Second 

Fig. 2. An illustration of a Rectified 50,000-Cycle Alternating 

Current. The dotted line indicates the average 

of the instantaneous peak values 

Now, if in a receiving circuit as described 
a continuous high-frequency current were 
induced in the aerial circuit, instead of in 
groups periodical^ by the spark trans- 
mitter, a direct current would flow through 
the telephone receiver pulsating at a fre- 
quency far too high to hear, the effect being 
identical to that of a continuous current. 

The alternating-current component neces- 
sary to reproduce speech may be obtained 
by varying the amplitude of this continuous 
current at the proper variable rate, which in 
turn can be accomplished by varying the rate 
of amplitude change in the high-frequency 
waves, intercepting and setting up corres- 
ponding currents in the receiving aerial. 

These various forms of currents can best 
be made clear by some simple diagrams 
illustrating the principles involved. 

Fig. 1 illustrates a direct current consisting 
of an alternating current of 1000 cycles fre- 
quency superimposed upon a continuous 
current. Such a frequency passing through 
a telephone receiver would produce a high 
pitched musical tone. 

Fig. 2 illustrates a rectified high-fre- 
quency current of 50,000 cycles, the negative 
half of the wave being suppressed by the 
rectifying device. Such a form of current 
passed through a telephone or direct-current 
instrument will give a response or indication, 
as if a continuous current were passing whose 
value is equal to the average of the instan- 

taneous values, or about 32 per cent of the 
peak value of the rectified wave. This is 
shown by dotted line. 

Fig. 3 represents a high-frequency current 
of 50,000 cycles, varying in amplitude so 
as to reach a minimum every 0.0005 of a 
second, or at a rate corresponding to 1000 
cycles per second. 

Such a current, passed through a rectifying 
device, is shown in Fig. 4 and the dotted line 
shows the equivalent average current which 
is of 1000 cycles, and produces the effect of 
such a current in a telephone receiver. A 
direct current added would make it identical 
to Fig. 1. A musical tone may thus be pro- 
duced in the receiving circuit by a periodic 
variation in the value of the current in the 
transmitting circuit. 

Under actual conditions the variations will 
follow an irregular curve, due to the over- 
tones and inflections of the voice, and the 
rectified high-frequency wave form will be 
complicated by the fact that the rectifying 
devices used do not rectify perfectly, -and 
because condensers are used to store the 
energy of succeeding waves so that the high- 
frequency current does not actually have to 
pass through the telephone windings. 

The usual form of radiotelegraphic receiv- 
ing apparatus is therefore suitable for tele- 

tiundred Thousandths 
of "Second 

Fig. 3. A Representation of a 50,000-Cycle Alternating 
Current of Varying Amplitude 

n n 


IS 20 

hundred Thousandths ofaSecand 

Fig. 4. The formation which the wave shown in Fig. 3 

assumes when rectified. The dotted line is of 

an equivalent 1000-Cycle Current 

phony, so that the modifications necessary are 
in the transmitting equipment. 

The first feature is that the transmitting 
station must be capable of generating high- 
frequency currents and radiating them so 



that the currents induced in the receiving 
apparatus when rectified will cause no dis- 
turbing noise in the telephone receiver. This 
may be done in two ways, either by a con- 
tinuous high-frequency wave, or by one gen- 
erated by the spark system described, the 
sparks, however, occurring one after the 
other so rapidly that their frequency is above 
the audible range where the telephone and 
ear are not sensitive and any resultant tone 
would not interfere with the reception of 

So far, the former method has proved the 
more practical, and the continuous high-fre- 
quency currents are generated either by an 
alternator of special design, or by some form 
of high-voltage direct-current arc shunted 
by a capacity and inductance. The Poulsen 
arc-generator is an apparatus of this type. 

The second important feature in the trans- 
mitter is some method by which the amplitude 
of the high-frequency current may be con- 
trolled and modulated by the voice so that 
the amplitude of the radiated waves follows 
closely every variation in the voice. Since 
the voice in speech is a complex set of sound 
waves varying continually in frequency and 
amplitude, and containing overtones, it will 
be realized that it is very difficult to modulate 
a current of, say, even 10 to 20 amperes 
through a wide variation and preserve at the 
same time the correct relative intensity of 
the different voice frequencies involved in 
order that the articulations at the receiving 
station is good. 

This matter of sufficient energy control 
is the one big problem in long-distance radio- 
telephony, and is the factor which has made 
it impossible so far to attain anything like 
the distance range that is accomplished in 

A great deal of work has been done by 
different investigators on the improvement 
of microphone transmitters which will handle 
heavy currents and give good articulations. 

The ordinary microphone transmitter, such 
as is in use on all telephones, operates with 
about a quarter of an ampere and about 10 
volts across it. This means a control in energy 
variation of but a few watts. Modifications 
in such a type of microphone transmitter 
may be made so that it will control several 
amperes, and special microphones have been 
built to handle considerably more current, 
but so far none have been perfected to control 
the large currents such as are used in high- 
powered radio-stations. 

There are two promising fields for radio - 
telephony. The first is for long distance, 
where wire telephony at present is impossible 
over submarine cables, and expensive on land. 
The other is for relatively short distance, for 
use between ships and from shore to nearby 
ships; the latter being used in connection 
with the land lines, so that conversation may 
be had with vessels not too far from land with 
the same ease that we now talk from one city 
to another. 

For the realization of this latter application 
several additional difficulties remain to be 
overcome. The great difference between 
transmitted energy and received energy pro- 
hibits the simultaneous use of sending and 
receiving apparatus, so that some form of 
throw-over device has been found necessary 
to change connections to either one or the 
other. Although the high-frequency alter- 
nator and the Poulsen arc give good results, 
both require more or less' attention and are 
not suitable for small ship installations. 

For use in connection with existing land 
lines, the problem of control is even more 
difficult, as here we have only minute cur- 
rents to effect the control of a large amount 
of energy. 

It is doubtful whether radiotelephony will 
ever supersede our present wire system on 
short distances over land, but it will un- 
doubtedly be of immense value in fields where 
the wire telephone is impracticable. 



By J. M. Drabelle 

The Iowa Railway & Light Company 

These note< describe the method adopted for supplying power to the Quaker Oats Company's plant after 
the growth in business had rendered the original plant too small for the present requirements. — Editor. 

A rather unusual and interesting installa- 
tion has been made by the Iowa Railway 
& Light Company of Cedar Rapids. Iowa, 
to supply electric power and high-pressure 
steam to the plant of the Quaker Oats Com- 
pany, which has at Cedar Rapids the largest 
cereal mill in the world. 

The Iowa Railway and Light Company 
generates power at 2300 volts, two-phase, 
(10 cycles; the Quaker Oats plant generated 
its power at 240 volts, three-phase, 60 cycles. 
The problem of arranging for the supply of 
purchased power may be stated as follows: A 
maximum of 3000 kw. at 85 per cent power- 

The power plant of the Quaker Oats Com- 
pany originally consisted of one 800-kv-a. 
engine-type alternator and one 400-kv-a. 
engine-type alternator. Two years ago, 
owing to the increased requirements for 
power, that company installed a 500-kw. 
85 per cent power-factor steam turbine and 
generator. The demand for power, however, 
kept increasing, and since the entire mill 
had been built around the engine room and 
boiler room, no space was available for the in- 
stallation of the needed additional machinery. 
The milling company therefore contracted 
with the local power and light company 

Fig. 1. 

The Method of Mounting the Potheads at the End 
of the Lead-sheathed Cables 

Fig. 2. Control Board, Showing Operating Lever of the 
Oil Switch and the Curve-drawing Instruments 

svo-phase, 2300 volts, was to be 

transmitted in underground cables a distance 
of 1S50 i< 'ansformer substation, that 

was located in the mill and there by means of 
the T conne< :ransformed to three- 

phase power Its. 

for electric power to be -used in driving the 
machinery, and for steam to be used in the 
cooking processes. 

The underground cable installation consists 
of three 750,000 cir. mil., two-conductor, 
concentric cables. Two of these cables are 


Fig. 3. 

Photograph Showing the Switchboard, and the 
Massive Busbars with their Supports 

maintained in service, and the third is held 
for a spare. One end of the cables, with their 
potheads, are shown in Fig. 1. An insulation 
of 5^-in. varnished cambric was used between 
the inner conductor and the outer conductor; 
a similar one being used between the outer 
conductor and the lead sheath. The lead 
sheath was }/% in. thick. 

Three water-cooled, 1500-kv-a., 2300 volts 
lu 1240 volts, two to three-phase transformers 
were supplied. Two are used in the T-con- 
nection and the third is held as a spare. The 
transformers are provided with round-pattern 
thermometers and with water bells. 

The switchboard installation, which is 
shown in Fig. 3, is particularly interesting 
and is easily the most unusual feature of this 
installation. The leads after leaving the pot- 
head are brought to an oil switch that is con- 
nected to the transformers. For protecting 
the cable from static disturbances, graded- 
shunt multigap arresters are provided. The 
low-tension leads of the transformer are 
brought through a single opening in the tank 
to prevent eddy currents; from the ter- 
minal board, 10-in. by 1 ^-in. copper bars run 
to a disconnecting switchboard that consists 
of six 6000-amp., single-pole lever switches. 
From this disconnecting switchboard, the 

Fig. 4. Photograph of the Indicating Recording 
Integrating Flow Meter used in 
this Installation 

Fig 5. Diagram of the Switchboard Wiring, Back View 



bus system is carried approximately 40 feet 
to the distributing board of the Quaker Oats 
Company. The busbars consist of six bars, 
two sets in parallel per phase, and each 
phase bar is made up of two 10-in. by 34 -in. 
copper bars. The weight of copper used in 
making these connections was 18,000 pounds. 

The control panel has an oil switch mounted 
on the back, and the following instruments 
are mounted on the front : One 3-phase, 
3-wire, 60-cycle, 240-volt, 10,000-amp. watt- 
hour meter; one curve drawing voltmeter, 
180-260 volts; and one curve drawing watt- 
meter, 4000 kilowatt scale. 

Calibrating links are provided for in order 
that testing instruments may be readily 
inserted in the circuits for checking. Two 

10,000-amp., 2000 to 1 ratio, current trans- 
formers operate the current coils of the 

Steam is supplied to the Quaker Oats 
Company through an 8-in. extra heavy pipe 
line. 1 150 feet long. All flanges, fittings, etc., 
are of cast steel. Expansion is allowed for 
by long bends. The steam pressure is 190 
pounds with 100 degrees superheat. The 
steam is metered by an indicating, recording, 
integrating steam flow meter, an illustration 
of which is shown in Fig. 4. 

This load of the Cereal company, which is 
being carried by the Iowa Railway & 
Light Company, is the largest power load in 
the state carried by a public service cor- 


By W. D. Bearce 

Railway and Traction Engineering Department, General Electric Company 

The semi-outdoor portable substation, consisting of an open section for the transformer, oil switch and 
other accessories, and two closed compartments for lightning arresters and the synchronous converter, is 
especiallj- suited for service on roads with limited overhead clearance. A small reduction in weight and total 
cost over the totally enclosed type is also possible, owing to the omission of a portion of the superstructure, 
and a further advantage lies in the fact that all high tension apparatus is kept on the outside of the cab. 
The following article describes in detail the equipment of a portable station of this type now being operated 
by the Berkshire Street Railwav. — Editor. 

In order to increase the flexibility of sub- 
station equipment many electric railways 
have adopted the expedient of equipping 
a portable substation which can be put in 
service at any point on the system on short 
notice. Temporary power requirements may 
occur at outlying amusement parks and fair 
grounds, or on extensions where permanent 
substations are under construction. An 
equipment of this kind can also be used in 
emergency as a reserve unit available at any 
of the permanent substations. 

During the past summer the Berkshire 
Street Railway placed in service a number of 
300-kw. commutating pole synchronous con- 
verters, one of which was installed in a port- 
able substation. The railway system on 
which this equipment operates includes about 
110 miles of interurban trackage in Western 
These lines radiate from 
Pittsfield. extending north to North Adams 
and Bennington, Vt. 

This portable substation is of the semi- 
outdoor type, consisting of an open section 
for the outdoor type transformer, oil switch, 

current transformer, choke coils, discon- 
necting switches, etc., and an enclosed sec- 
tion divided into two compartments for the 
lightning arrester and synchronous con- 
verter with switchboard equipment. Owing 
to the low clearances of several overhead 
bridges the car height has been limited to 
1 1 feet 6 inches above the rails. 

The car body is an all-steel structure built 
in accordance with Master Car Builders' 
standards and fitted with steps, hand rails, 
ladders etc., as required by the Interstate 
Commerce Commission under the safety 
appliance acts. 

The under frame is made up of four 12- 
inch steel channels extending the entire 
length of the platform. The two center 
channels are tied together by three-eighth- 
inch top and bottom plates, forming a box 
girder, thus securing the. necessary stiffness 
without depending upon the car flooring. 
All of the channels are securely fastened 
together at the ends by steel members and 
cross-braced by 12-inch steel channels. ■ 
There is also a cross bracing of six-inch steel 



angles in the intervening central section of 
the framing. 

A suitable foundation for the converter 
is provided by two pairs of six-inch steel 
channels placed across the car and firmly 
riveted to the underframing. The space 
between each pair is filled with concrete. 
Provision is also made for the usual levelling 
plates and anchor bolts. Ventilating open- 
ings in the floor of the car insure a supply of 
cool air when the machine is in operation. 
These openings are fitted with removable 
sheet iron covers and permanent wire mesh 

The car flooring consists of one-quarter- 
inch sheet steel, which extends across the 

formed by thin sheet steel framed with angle 
iron. This shield protects the high tension 
bushings of the transformer which are 
brought out in a horizontal direction on 
account of the very limited overhead clear- 
ance. Some protection is also afforded tin- 
oil switch units and the current transformer. 
At the other end of the supporting frame, a 
short cover extends over the incoming line 
leads and those to the lightning arrester 
compartment. The incoming insulators at 
each side of the choke coils are suspended 
from cross steel angles tied into the framing. 
Doors are provided on each side of the 
closed' compartment and there are two glass 
windows located in each side of the operating 

Fig. 1. Side View of the Semi-outdoor Portable Substation 

plates on top of the center girder. The side 
and roof framing of the closed section and 
partitions of the car consist of steel channels 
and angles suitably braced and riveted to- 
gether. The sides and roof are enclosed with 
sheet steel. A section of the roof over the 
converter is fastened with bolts so that it 
may be readily removed for installing or dis- 
mantling the apparatus when a crane is 
available. A galvanized sheet metal ceiling 
is built on the interior, forming air pockets 
which prevent any direct radiation of heat 
when the car is standing in the sun and also 
to drain any condensation to one side of the 
car away from the apparatus. 

At the open end of the car a framework 
of channels is erected which forms a support 
for the disconnecting switches and choke 
coils. The framework, together with the 
transformer, also supports a snow shield 

compartment. The windows are pivoted at 
the center to allow for suitable ventilation. 

The trucks are of the diamond frame, arch 
bar type equipped with 33-inch wheels 
mounted on Master Car Builders' standard 
steel axles and fitted with cast iron journal 
boxes. They are designed to take a curve of 
approximately 40 foot radius. 

Standard automatic air brake equipment is 
supplied with shoes acting on all wheels and 
with hose connections, thus conforming to 
the Master Car Builders' standards for steam 
railroad lines. A handwheel and brake shaft 
is also provided for operating the brakes at 
one end of the car. Current is taken into the 
car through three 33,000-volt disconnecting 
switches designed for outdoor service. These 
switches are so connected that they cut out 
all of the apparatus in the station, including 
the lightning arrester. An eight-foot switch 



Fig. 2. The 300-kw., 600-volt. Commutating-pole 

Synchronous Converter Installed in the 

Portable Substation 

hook is furnished for manual operation of the 
switches. From these switches current passes 
through three 200-ampere choke coils sup- 
ported in a horizontal position and thence 
to the 300-ampere. 45,000-volt, triple-pole oil 
switch. This switch is enclosed in three 
separate tanks, the mechanism being operated 
from a single handle mounted on the control 
panel. A separate current transformer is 
provided for automatically tripping the oil 
switch in case of overload and for operating 
a bell alarm to notify the attendant when the 
switch opens. The oil switch is instantaneous 
in action, thus affording complete protection 
to machines and feeders under short circuit 

The transformer is an oil insulated, self- 
cooled, outdoor-type unit, rated 330-kv-a., 
three-phase, 25 cycle- Voltage taps are 
arranged on the primarv side for operating 
13,000 or 11,000 volts Y, by using 
either series or multiple connection of the 
primary coils. The substation can thus be 
connected to any of the high tension lines on 
the company's system. The secondary wind- 
ing is designed for 385 volts and has 50 per 
cent -aps. The secondary lea* 1 

enclosed in a sheet iron box from which con- 
nections are made through a conduit to the 
operating compartment. 

The transformer is of the standard railway 
type, designed with high inherent reactance 
and giving a practically flat compounding 
on the d-c. side of the converter. The con- 
tinuous rating is in accordance with the 
A.I.E.E. recommendations, the allowable 
temperature rise after 24 hours' operation 
at full load being 35 degrees C. The tem- 
perature rise after two hours' operation at 150 
per cent load will not exceed 55 degrees C. 

The operating compartment of the car 
contains a three-phase, 600-volt commutating 
pole synchronous converter operating at 
750 r.p.m. and a three-panel controlling 
switchboard. The converter has a normal 
rating of 300 kw. continuously, standard 50 
per cent overload for two hours, and a 
momentary capacity of three times normal, 
or 900 kw. The machine is started from the 
a-c. end from 50 per cent starting taps on the 
transformer. There is also a series resistance 
in circuit to cut down the initial rush of 
current. The temperature guarantees and 
insulation tests follow the recommendations 

Fig. 3. 

The Control. Feeder and Starting Panels left to right' 
of the Portable Substation Switchboard 



of the A.I.E.E. The compound field is de- 
signed to give a practically flat compounding 
at all loads without shunt field adjustment. 
The machine is equipped with speed limiting 
device, mechanical end play, field break-up 
switch, equalizer switch, negative line switch 
and shunt field rheostats. There is also the 
usual brush raising device to be used with 
a-c. starting. An opening is provided in the 
floor for connecting the equalizer to the 
stationary substation, if parallel operation is 

The switchboard is of natural black slate 
mounted on pipe framework and includes a 
transformer panel, d-c. feeder panel, and an 
a-c. starting panel. The transformer panel 
is 4S inches by 20 inches and the feeder and 
starting panel 48 inches by 16 inches. All 
the panels have 20-inch sub-bases. The 
transformer panel carries a 1500-ampere 
ammeter with shunt, a 750-volt voltmeter, 
a 300-0-300 scale wattless component indica- 
tor, two two-point potential receptacles, and 
the operating lever for the automatic high 
tension oil switch. 

The d-c. feeder panel is equipped with a 
single-pole, 600-volt, 1000-ampere carbon 
break circuit breaker which is hand-operated 
and has a bell alarm switch; a back-of -board 
mounted rheostat with handwheel for the 
converter field; a single-pole, GOO-volt, 1000- 
ampere line switch; and a 600-volt, 500- 
ampere, two-wire recording watthour meter 
mounted on the sub-base. 

On the synchronous converter starting 
panel there is a double-pole, double-throw, 
SOO-ampere starting switch and two double- 
pole, single-throw, 100-ampere switches with 
enclosed fuses for lighting and heating circuits. 

The d-c. lightning arrester for the 600-volt 
feeder circuit is of the aluminum cell type and 
is mounted in the rear of the switchboard 
panel. The high-tension multigap lightning 
arrester is contained in an enclosed central 
room and consists of a series of spark gaps 
shunted by graded resistances, but without 
series resistance. It may be connected for 
protection of either the 33,000-volt circuits 

or the 13,000- and 11,000-volt high tension 

The current for lighting and heating is 
taken from the partial voltage taps on the 
transformer secondaries. The heaters are 
fastened to the partition and side of the car 

. 4. The Outdoor Section of the Portable Substation 
Showing Transformer. Oil Switch, Choke Coils 
and Current Transformer 

at one end of the switchboard. They con- 
sist of three units normally rated at 900 watts 
each. Separate switches are provided to 
secure a gradation of heat. 

The principal dimensions and data of this 
substation are as follows: 

Length overall 38 ft. 

Width over sides of car S ft. 4 in. 

Maximum width (over side channels! s ft. 6 in. 
Heightoverall (includingrunningboard ll 1 ft. 6 in. 

Height of floor above rails 3 ft. 8}^ in. 

Total length of enclosed cab. 23 ft. Li in. 

Length of converter or operating room . . 14 ft. 6 in. 
Length of lightning arrester compart- 
ment . 9 ft. 

Length of outdoor section 14 ft. 6 in. 

Truck base 25 ft. 

Wheel base 5 ft. 2 in. 

Wheels 33 in. 

Track ■ i tandard 4 ft. 8% in. 

Total : 80,0U0 lb. 




By Robert Reid 

Mechanical Superintendent's Office, General Electric Company 

The author of the article below first briefly reviews both the method of impregnating coils with an 
insulating compound and the equipment necessary for the process. Following this is a description of the 
design and the recent installation of a plant which is capable of impregnating nine-foot by twenty-foot 
coils. — Editor. 

The proper impregnation of the various 
types of armature coils, field coils, relay coils 
and coils used in industrial control work, 
also the drying and filling of wood is a subject 
which is of vital interest to all electrical 
engineers. How best to secure the complete 
penetration and filling of all the interstices 
of the coils with an insulating compound 
that will not be too brittle when cold, and 
yet capable of withstanding a reasonably 
high degree of temperature, is the problem. 
While undoubtedly all engineers are more or 
less familiar with the general process of 
impregnation, the following facts in con- 
nection with the apparatus may be of interest. 

The installation consists principally of two 
tanks; one is known as the mixing tank in 
which the compound is 
melted and thoroughly 
mixed by being stirred with 
paddles on a vertical shaft, 
the other is known as the 
treating tank in which the 
articles to be impregnated 
are placed. The two tanks 
are connected at the bottom 
by means of a pipe with a 
shutoff valve. Suitable 

When the proper period of time for complete 
exhaustion has elapsed, the valve in the pipe 
connecting the two tanks is opened thus 
allowing the compound to flow into the 
treating tank. The temperature of the 
treating tank in the meantime has been 
maintained at such a point as to insure the 
thorough drying of the articles and also to 
prevent any deterioration of the material. 

In some cases, when sufficient impregnating 
material has entered, the valve is closed and 
air compressed to about 100 or 12.5 lb. per 
sq. in. is admitted directly on top of the 
compound. This pressure is maintained for 
about one or one and one-half hours. This 
method gives good results when the coils are 
short and do not lie near the surface of the 

vacuum pumps, air com- 
pressors, condensers, and air 
dryers are also essential 
parts of the outfit. 

A description of the oper- 
ation of impregnating fol- 

The treating tank having 
been filled with the coils, 
etc., the cover is placed and 
securely fastened with 
heavy swing bolts and nuts. 
The vacuum pump is then 
started and a vacuum as near as possi- 
ble to 29 in. or 30 in. is maintained 
for from one to one and one-half hours. 
During this time all the air from the 
interior parts of the coil and wrappings is 


&re mm~2Z&P*<* 




|^7^^^^ ftttwe 





wf Mj^i < !J5bE Bk"^ 

Fig. 1. Lower Forms for Impregnating Pit Walls set in place 

compound, otherwise there might be great 
danger of some portion being exposed. 

The other method is to pass all of the 
compound over from the mixing to the treat- 
ing tank and then turn the air pressure into 
the mixing tank. This method ensures the 



treating tank being always full during the 
operation, and eliminates all danger of any 
parts being exposed during the time that the 
coils are under pressure (in this method also 
the time is usually equal to about one or 
one and one-half hours). The compound is 
then returned by opening a valve in the 
mixing tank to the atmosphere and introduc- 
ing compressed air to the treating tank, 
which forces the compound back. 

The air dryer, as the name implies, is for 
drying or removing as much moisture as 
possible from the compressed air before 
allowing it to enter the tanks, while the 
condenser takes care of the air on the vacuum 

The impregnating tanks are usually made 
double, that is, with an inner and outer wall. 
Some of the smaller sizes are made of cast iron, 
but all of the larger sizes are of steel plate 
with riveted or welded joints as may be 

The chamber between the two shells is 
for the heating element; steam at such pres- 
sure as will yield the required temperature 
being used direct. In some cases, in the 
larger tanks, a steam coil submerged in oil 
is located in this space. The body of hot oil 

Fig. 2. Loads as placed on the Tank to Facilitate Sinking it into place 

tends to maintain a more uniform temperature 
in the tanks. 

There are other designs consisting of only 
a single tank with a steam coil inside it. 
In such a tank the heating coil comes into 
direct contact with the compound. All tanks 

regardless of the type are covered with 
asbestos or magnesia covering. 

In one of the latest impregnating plant 
installations, the tanks have a double shell 
and the space between contains only hot oil 
as the heating medium. This appears to have 
many points of interest and advantage. 
When steam-heating coils are used the pres- 
sure must be maintained at all times because 
if. for any reason, the pressure drops there is 
a corresponding drop in the temperature 
of the compound. In the case of direct 
healing by oil, if the body of the oil is 
sufficiently large, a more uniform condition 
can be maintained. The temperature is 
not easily affected ; it will decrease very slowly, 
and yet is capable of being raised in a 
very short space of time to the required tem- 

The great length of armature coils for 
horizontal turbine-generators was largely 
responsible for the installation of this latest 
and also one of the largest sets of its kind in 
existence. The inner shell of each tank is 
9 ft. inside diameter by 20 ft. deep in the 
straight part while the outer shell is 10 ft. 
in diameter, which gives about a 4 in. space 
between the two shells. The lower heads are 
of dished steel, rolled to 
shape. The upper part of 
the shells are securely riv- 
eted to a heavy cast steel 
flange that carries forty- 
eight 2 in. eyebolts by 
which the cover is fastened 
down. The cover also is 
made of a dished steel head 
with a cast steel flange. 

Both mixing and treating 
tanks are similar in con- 
struction except for a slight 
difference in the covers. The 
mixing tank is provided 
with a stirring device and 
its cover carries the neces- 
sary gearing, bearings, mo- 
tors, etc., for this operation. 
Oil having been chosen 
after careful consideration 
as the heating medium, it 
became necessary to con- 
sider its heating and proper 
circulation. Each tank has its own heater, 
circulating pump, piping system, and tem- 
perature recording gauges and thermometers, 
so that it is possible to see at all times 
just what temperatures have been obtained, 
both in the oil and in the compound. 



Fig. 3. View of the Tank in place within the Pit 

horse power on the compressor to 5 horse 
power on the oil pumps. 

When it is considered that these tanks 
must be of sufficient strength to withstand 
both a collapsing and a bursting effect it can 
readily be seen that the material and work- 
manship must be of the best. Roughly, a 
vacuum of 30 in. will give a collapsing 
pressure of about one ton per sq. ft. of surface. 
This means that the cover of the tank must 
sustain approximately a total load of 65 
tons while under vacuum. 

It may be of interest to know that a total 
of about 3600 gallons of heating oil are 
required for both tanks, while 220 barrels 
of compound are required to fill the mixing 
tank. Both tanks, and all of the piping and 
valves are very carefully covered with 
magnesia covering so that the difference of 
temperature of the heating oil entering and 
leaving the heater is very small (about 5 deg. 

The installation of this plant presented 
many difficulties. Only about 4 ft. of the tanks 
project above the floor level, the remainder 
being below in a basement. This basement 
(the outside measurements of which are 
4(1 ft. by 26 ft. by 25 ft. deep), made of 

Hot oil leaving the lower 
part of the heater passes to 
the bottom of the large 
tanks, thence around the 
bottom and up the sides 
leaving at the top, going 
thence through the circu- 
lating pump to the top <,;' 
the heater, etc. 

While both tanks have 
an independent circulating 
or heating system, they are 
both connected to a com- 
mon expansion tank that 
placed at the highest point 
on the line. The top of 
this tank is sealed to pre- 
vent oxidation as far as pos- 
sible, and has an overfly 
connecting with the sewer 
so that any unforeseen ex- 
pansion of the heated oil 
Vie taken care of. 

All machinery (air com- 
vacuum pump, stir- 
ring pad r hoist for 

ting tanks, and oil circulating pumps), 

connected with the apparati motor 

n, the m arving in size from :!."> 

Fig. 4. General Interior View of the Impregnating Plant. The Tops of the Two 
Nine-foot by Twenty-foot Tanks are shown in the center 

concrete, is very strongly reinforced, and is 
damp proof. Due to its nearness to a massive 
building, and to the nature of the soil, it was 



decided to make only a partial excavation and 
to construct this enormous concrete tank in 
the excavation, and by removing the soil 
underneath, allow it to 
gradually settle to the re- 
quired depth. It was first 
attempted to remove the 
soil by washing it out with 
water under pressure but 
the air-lock system was 
finally adopted. Two air 
locks were used and the 
concrete tank was success- 
fully lowered to place. (A 
number of branches of trees 
were encountered at a depth 
of about 25 to 30 ft., the 
wood resembling that of the 
elm tree. It was very light 
in color when broken but 
rapidly turned dark when 
exposed to sun-light). 

Fig. 1 shows the setting up 
of the forms for the concrete. 

Fig. 2 shows how the tank 
was loaded down to assist 
in lowering it into place, 
and Fig. 3 shows the tank 
in place and the workmen 
removing the air-locks pre- 
paratory to closing the openings in the floor. 
Figs. 4 and 5 show views of the completed 

interior. Fig. li shows a smaller outfit of 
which the impregnating tanks are 3 ft. in 
diameter and 4 ft. deep. The heating element 

Fig. 5. 

The Air-Compressor. Vacuum Pump, and Condenser installed for use 
in the Impregnating Plant 

in these tanks is steam introduced directly into 
the space between the inner and outer shells. 

Fig. 6. A Three-foot by Four-foot Impregnating Set of Tanks at the Left and a Four-foot by 
Eight-foot Synthetic Resin Vulcanizer at the Right 



By George P. Roux 
Consulting Electrical Engineer, Philadelphia, Pa. 

The merits of the scheme of employing the open-delta or V-connection of transformers in a case of 
emergency or as a permanent condition has often been questioned. The proposition has both been opposed 
and defended. The following article presents a clear analysis of the matter and after considering the scheme 
from the standpoints of capacity, stresses, stability, etc., conclusions are drawn which are formulated at the 
end of the article. — Editor. 

A style of connection of single-phase 
transformers for three-phase transformation of 
voltage which is very often used, not only in 
case of emergency but also in ordinary opera- 
tion, is the open-delta or 1 '-connection. 

For ordinary and permanent operation, the 
use of two single-phase transformers V con- 




line B 







Fig. 1. Diagram of Three Single-Phase, 10 kv.a., Trans- 
formers Connected in Closed Delta 

nected to transform a three-phase current 
to a lower voltage, appeals to many of us, 
especially in cases where an increase in load 
is expected in the future, when a third trans- 
former can be added changing the connec- 
tions from open to closed delta. The initial 
investment in transformers is reduced one 
third, while provision is made for the future 

As an emergency connection of two trans- 
formers in a bank of three, for example, in 
case of an accident to the third one, this 
style of connection is very often resorted to, 
with acceptable results, while the disabled 
transformer is being repaired. 

In both cases, however, we are apt to lose 
sight of the fact that the capacity of two trans- 
formers connected in open delta is not equal 
to the sum of the capacity of each one, and 
furthermore, that a number of conditions are 
changed which affect the operation of both 
primary and secondary circuits, as well as the 
operating characteristics of the transformers 
themselves. These conditions we propose to 
review and analyze in this article. 

In order to have a clear view of the situa- 
tion, let us take first the case of three single- 
phase transformers connected in delta, each 

transformer having a capacity of 10 kv-a., 
100 volts and 100 amperes, as shown in Fig. 1 ; 
and deal in all the following cases with trans- 
formers of the same ratio, same impedance, 
and connected to a non-inductive load with 
each phase balanced. We have then at full 
load a voltage per line of 100, and a current 
per line of 100X2 cos. 60 deg. = 100 X 1.73 = 
173 amperes. Since the phase-angle relation 
between the transformers is 120 deg., or twice 
60 deg., the relation between the line current 
and the phase current is therefore 30 deg. for 
one branch, and 60 deg. between two 

Adding vectorially the two currents with a 
phase angle of 60 deg., as in Fig. 2, OA +OB = 
OC, the line current. Also (CM +05) cos. 
30 deg. =line current OC. 

As each line is in parallel with each phase, 
the phase voltage is equal to the line voltage 
or 100 volts. Thus, the total power in the 
bank of transformers as shown and con- 
nected in Fig. 1 is 3X100X100=30 kv-a. or 
V3 X 100 X 173 = 30 kv-a. 

Assume, for some reason, that transformer 
77/ must be removed from the delta-con- 
nected bank, leaving only transformers / and 
II in sendee, as for instance in case of a 
breakdown in transformer 777. 

Fig. 2. 

Vector Addition of Two Currents 60 Electrical 
Degrees Apart 

We now have a bank of only two trans- 
formers which are connected as in Fig. 3, 
and each is of the same size and capacity 
as those in Fig. 1. 

It is obvious that in a three-phase system, 
in order to supply 30 kv-a. from two trans- 
formers instead of three, the line current and 



voltage must be the same in each case, that is, 
100 volts and 173 amperes. 

The capacity of each transformer in a bank 
of three, as in Fig. 1 connected in closed 
delta, was 

V3X 100X173 ,_. 

— = 10 kv-a. 


The capacity of each transformer in a bank 
of two, as in Fig. 3 connected open-delta 
or V, would appear to be 

V3 X 100X173 1E1 

— - — = 15 kv-a. 

We see at once that the two remaining 
transformers, 7 and 77, in this new system of 
connection will be loaded above their rated 
capacity, which is 10 kv-a. only; and, although 
it may seem that the increase of current in the 
winding of each transformer is only 50 X 33^3 
= 16% per cent, such is not the case. 

Analyzing the conditions set forth in Fig. 3 
as to phase relation of the current in each 
transformer, we see conditions peculiar to 
two transformers operating in series, viz. 
across lines A and C. 

First: The current in line A (173 amperes), 
entering transformer 77 at a must neces- 
sarily flow with all its intensity in the wind- 
ing, and therefore the value of the current in 
transformer II is 173 amperes, it being raised 
from 100 amperes in the closed delta connec- 
tion to this 173 in the open-delta connection. 

This is due to the fact that since the total 
power to be supplied is equal to EI \/^~ t 
where E is the line voltage and I the line cur- 
rent, each transformer in the open-delta con- 

nection has to supply — - — : and each 

winding will be subjected to a voltage E, 
(line voltage), and in addition to a current 

Line A 









Fig. 3. Diagram of Two Single-phase, 10 kv-a., 
Transformers Connected in Open Delta or V 

equal to 7, (line current). Such a condition 
gives rise to a phase displacement, between 
the line voltage E and the current in the 
transformer winding, equal to: 


~ =0.866 = cosine 30 deg. 

In the delta connection, we note that they 
were connected with a phase relation of 120 
deg., or 180 deg. — 120 deg. =60 deg. between 
currents in the windings. 

In the open delta this condition has been 
changed and the difference in phase between 
the current in each winding is 180 deg.- 

Fig. 4. Vector Addition of Two Currents 
120 Electrical Degrees Apart 

(2X30 deg.) = 120 deg. The resultant of 
two currents 120 deg. apart has the same 
value as either of the components as shown 
vectorially in Fig. 4, where OA plus OB = OC, 
as also OA cos. 120 deg. = 0.4 because cosine 
120 deg. =1. 

Therefore the line current 7=173 amperes 
has a value of 7 cos. 120 deg. = 173X1 = 173 
amperes in the winding of transformer 77: 
and, as transformer 7 operates under identical 
conditions, the current flowing in it is of 
equal value. 

In Fig. 5 we have represented the three 
transformers connected in delta. The imme- 
diate effect resulting from the release of trans- 
former 777 is similar to the case of a step 
ladder in which the braces, which keep the 
legs from spreading, have been severed owing 
to the weight on the ladder and to other con- 
current conditions; the legs have spread apart, 
each one to an angle greater than the former 
by 30 deg., therefore the stress in each leg 
has increased to a value equal to 

= 5. 

cos. 30 deg. 

Fig. 5. Diagram of a Closed-Delta Connection 
Made up of Three Single-phase Transformers 

We can realize that this new condition 
is very unstable, as there is nothing to prevent 
the legs from spreading further apart, on 
account of the absence of the connecting 
member 777, which, to a certain extent 



and within a certain limit, kept the two legs 
/ and II in a determined angular position. 

We can then write again, but this time 
correctly, that the output of the two trans- 
formers left in open delta, from the previous 
bank of three in closed delta, is 

Fig. 6. A V-Connection Constructed from the Delta of 

Fig. 5 by Disengaging Transformer III and Spreading 

the Phase Angle Between I and II 

El \ 3 


34.04 kv-a., 

cos. 30 deg. 0.866 

and each transformer must have a capacity i if 

3464 I7Q01 

— — — = 1 , kv-a. 

It follows that when two transformers are 
left in open delta from a bank of three in closed 
delta, to supply under the same conditions 
the full load of the three delta-connected 
transformers, they will each be subjected to 
an overload of 73.2 per cent, or, vice versa, 
the line can be loaded to only 57.7 per cent 
of its rated supply capacity to operate the 
transformers under normal rating conditions. 

The voltage and current relation in the 
two systems of connections is best shown in 
Figs. 7 and 8, from which it can be seen that 
while the three delta-connected transformers 
operate with current and voltage in phase in 
their winding, in the open delta they operate 
a1 86.6 per cent power-factor, or specificallv. 
with a current lagging behind the voltage 
30 deg. in one transformer and leading by 
30 deg. in the other. 

Taking for instance the current in line C 
of the delta-connected transformers in Fig. 7, 
which is in phase with the imaginary Y cur- 
rent of the transformers, we can see that upon 
reaching the point of connection of transform- 
ers II and III the line current splits itself 
into two components each 30 deg. apart 
from the line current or in proportion to the 
impedance of the paths offered to its flow. 
which in our case is the same, each component 

being equal to '—- = 100 amperes. If 

2 cos. -in deg 

ance of each transformer is not the 

-ante, as in the case of unbalanced load, then 

urrent will divide in the inverse ratio to 

dance. We note that each current 

component is here in phase with the voltage. 

Looking now at the open-delta connection 
of transformers I and 77 in Fig. S, and taking 
the line C again, the line current is, as in the 
above case, in phase with the Y voltage of the 
transformers; and when it reaches trans- 
former II it finds only one path to follow, 
instead of two as before, and strikes trans- 
former II at an angle of 30 deg. behind the 
voltage, thus lagging 30 deg. or 0.S66 per 

In line B similar action takes place (except 
that line current, B, leads voltage of trans- 
former I) and the two currents emerging 
from transformers / and II combine with a 
phase angle of 120 deg. into a resultant equal 
to each of its components, as explained 
before and shown in Fig. 4, or 173 amperes 
which flows back to the generator through 
line .4. 

This operation is repeated in each cycle 
and in each phase, and needs no further 

There are other peculiarities inherent to 
the open-delta connection which materially 
affect the operation of three-phase trans- 
formers or apparatus so connected. 

Fig. 7. Diagram of the Phase Relations of the Voltages and 
Currents in a Closed-Delta Transformer Connection 

The line current entering a system of inter- 
connected transformers divides itself in the 
inverse ratio to the impedance of the paths 
offered for its flow. Any difference in the 
value of the respective internal impedances 
is likely to cause an unbalance of the sec- 
ondary voltage and primary current, which 



under certain operating conditions very often 
met may reach dangerous proportions. 

Likewise, the production of electrostatic 
stresses has fatal consequences, due to the 
fact that the mean potential of the primary 
windings is not the neutral potential. 

For these reasons, the use of the open- 
delta connection with transformers of high 
voltage is not recommended, as destructive 
potentials may be caused by unbalanced 
loads with electrostatic stresses, due to the 
instability of the internal impedance of the 
transformers under operating conditions which 
in turn cause an unbalanced voltage in the 

For low primary voltages, 10,000 volts or 
less, and relatively small installations, this 
system can be used sometimes to decided 

At non-inductive load, it will be found that 
the current is lagging in one transformer, and 
leading in the other. When the load becomes 

Fig. 8. Diagram of the Phase Relations of the Voltages and 
Currents in an Open-Delta Transformer Connection 

inductive, the relation of current to voltage 
changes and the phase angle displacements 
increase in one transformer and decrease in the 
other, thus making very unstable conditions. 
Therefore, if, when supplying a balanced 

three-phase load it is found necessary to con- 
nect a single-phase non-inductive circuit to 
the open-delta connected transformers, it is 
advisable to take this circuit off the leads a 
and c of the bank (see Fig. 3) rather than off 
one or the other transformer leads. 

Concerning the reduction in the transformer 
bank capacity, it is worth noting that in 
three-phase to three-phase transformation the 
three styles of connection commonly used also 
give different ratings. 

Taking single-phase transformers provided 
with all necessary taps and connecting them 
in different ways, their rating taken as a 
bank compares as follows: 

Three transformers in closed delta — Total 
capacity 100.00 per cent. 

Two transformers in open delta — Total 
capacity 57.7 per cent. 

Two transformers in T or Scott connection 
■ — Total capacity 62.2 per cent. 

In figuring the proper capacity of trans- 
formers which are found advisable to operate 
in open delta to supply a certain load, the 
size of each transformer in kv-a. is found by 
dividing the total load to be supplied by 2 
and adding 15.466 per cent to the result, in 
order to supply the kv-a. due to the lagging 
current peculiar to this type of connection. 

From the preceding discussion, which is 
based on permissible heating, it appears 
that the T, double- T, or Scott connection of 
two transformers for three-phase operation 
is more economical than the open-delta 
connection, on account of the somewhat 
greater available kv-a. capacity. 

However, since the open-delta connection 
uses no taps, it permits a greater simplic- 
ity in the construction of a transformer 
which naturally lowers the cost of the unit 
and renders attainable a greater degree of 
ruggedness for the same cost than can be 
obtained in a transformer for T connection 
which requires taps. Transformers without 
taps are better balanced internally to with- 
stand electromagnetic and electrostatic 

In conclusion, it may be said that there may 
be cases, however, where transformers having 
special taps for the T connection might well 
be found useful in times of emergency or m 
cases of temporary installation, because of 
their greater kv-a. capacity as brought out in 
this article. 




Part IV (Xos. 19 to 2s inc.) 

By E. C. Parha.m 
Constriction Department, General Electric Company 


A common symptom of disorder in direct- 
current generators and motors is the blacken- 
ing of the commutator bars and the ultimate 
eating out of the mica from between them, if 
the cause of the symptom continues. The 
simplest case to diagnose is where the black- 
ening is local to two or more pairs of bars 
that are so located at intervals around the 
commutator that the locations seem to be 
associated with the plan of winding and con- 
necting the armature. A poor soldering job 
throughout will cause so many poor con- 
nections that the commutator will blacken 
all over. An actual open circuit due to a 
coil being mechanically injured, or to its 
having burned in two, or to one of its leads 
having burned free from the commutator, 
will cause a traveling spark that will soon 
bum the mica from between the affected bars 
and characterize the trouble at once. 

A certain motor had been driving shop 
shafting for several years, during which time 
the commutator held a high polish. The 
motor was then transferred to duty in which 
the armature was reversed frequently while 
under load, under which conditions it soon 
developed armature open-circuit symptoms. 
Resoldering of the leads brought about 
no permanent improvement. The brushes 
sparked badly, but the machine was main- 
tained in service until a new armature could 
be obtained. The new armature performed 
its functions perfectly. 

It was decided then to rewind the old one 
because its insulation had become too dry. 
On stripping, several coil leads were found to 
be actually broken and several more were 
about to break. Further inspection dis- 
closed the fact that the commutator was so 
loose that it could be worked back and forth 
one-quarter of an inch. (The armature had 
not been designed for reversal operation and 
consequently it could not withstand the 
i at service. 


Plau ill mptoms may sometimes prove 
misleading. An operator was using a three- 

phase induction motor to run a laundry 
machine. The motor had worked satis- 
factorily for two years and then began to give 
starting trouble. Sometimes the closing of 
the compensator would start the motor and 
sometimes it would not. At such times as 
the rotor failed to move it would hum; any- 
one familiar with the symptom would have 
immediately suspected single-phase operation, 
hence a loose or open connection. In all the 
preceding cases the last of several attempts 
to start the motor had been successful. 

One day the motor could not be started 
at all ; the operator opened the compensator 
and cleaned the contacts; the motor then 
started promptly and ran without trouble all 
the forenoon. The next afternoon the motor 
again could not be started, and the cleaning 
of the compensator contacts this time did 
no good. While the operator was throwing 
the compensator "on" and "off," trying to 
humor the motor into a start as he had done 
many times before, a workman threw a 
rolled-up wet shirt at another workman. 
The shirt hit the ceiling and then dropped 
onto the stator wires just where they come 
out of the motor. The motor started at 
once and after throwing the compensator 
over normal operation continued. 

The operator was quick to connect the 
throwing of the shirt and the starting of the 
motor; and he had a mind to leave the shirt 
where it was as a permanent institution but 
more mature thought prevailed and he found 
that the pushing in of the stator wires was as 
effective as the shirt. The trouble was a 
stator lead which was loose inside the sleeve 
connector, the two being held together by 
the tape that bound the joint. Sometimes 
the two made contact and at other times they 
did not, hence the erratic starting actions. 


When an alternating-eurrent generator 
supplies under-loaded transformers and in- 
duction motors, the current of the generator 
lags behind its e.m.f. Under this condition 
more exciting current must be furnished by 
the exciter if the alternating voltage is to be 



maintained at normal. (The effect of lagging 
armature current is to cause the magnetizing 
action of a given armature coil to oppose the 
magnetizing action of that field pole which 
it is opposite. This decreases the resultant 
field cut by the armature conductors so that 
they do not generate as high an e.m.f. as they 
do when the voltage and current are in phase 
and the opposing armature reaction is at a 
minimum. In other words, a lagging current 
tends to demagnetize the field poles, and. if 
normal voltage is to be maintained, the 
field strength must be restored by increas- 
ing the amount of current drawn from the 

An operator complained that his exciter 
commutated badly and that he was unable 
to keep up the voltage on the alternator. 
These two symptoms in themselves suggested 
that an overload was the trouble and excessive 
heating of the exciter confirmed this suspicion. 
Since the exciter circuit included no ammeter 
its output was not shown. An inspector cut 
in an ammeter which indicated the exciter 
to be continuously overloaded 40 per cent. 
The operator then stated that the exciter was 
guaranteed to maintain normal a-c. voltage at 
0.8 power-factor and that his power-factor 
was better than 0.8. Rough calculations 
based on station wattmeter, ammeter, and 
voltmeter readings showed that the power- 
factor was between 0.55 and 0.60 at the time 
of the readings. The operator was well 
enough satisfied with these results to sub- 
stitute a larger exciter for the work. 


If trouble of any kind occurs in lay-outs 
that involve electrical apparatus it is usually 
the case that some one of the electrical 
devices is promptly blamed as being the cause. 
Apparently it never occurs to many operators 
that troubles may arise from abnormal con- 
ditions in the connected load. 

In a certain instance an operator com- 
plained to the power company that the volt- 
age supplied was varying widely and was 
causing speed variations in an induction 
motor that was driving a centrifugal pump 
The power company failed to see how this 
could be possible but, to satisfy the con- 
sumer, applied a recording voltmeter, which 
showed the voltage to be well maintained. 

On looking further for the trouble, it was 
found to be due to a hot box on the pump 
shaft; the box would alternately bind and 
release, thereby causing the motor speed to 
vary according to this variable load which 

was imposed upon its regular load. (The 
operator should have observed that voltage 
variations could not have been the cause of 
trouble because other induction motors on the 
same service were not affected.) 

A short time afterward, the same pumping 
unit, after working all day, refused to start 
on closing the control switch the next morn- 
ing. The panel starting-contactor closed 
promptly but the operator reported the motor 
was "dead." It took an inspector about five 
minutes to find out that there was nothing 
wrong with the motor. On turning the rotor 
back to let out the back-lash due to the belt 
coupling, and closing the control switch, the 
rotor promptly took up the back-lash but 
refused to rotate further. An investigation 
showed the trouble to be in the same pump 
bearing that had given trouble before; it was 
now frozen tight. A new lining, plenty of 
oil, and a smooth shaft remedied matters. 


A solid iron core would offer such a low 
resistance to the eddy currents set up by the 
core cutting the magnetic lines of the field that 
these currents would be very large and cause 
prohibitive heating. Laminations, when in- 
sulated from each other, introduce resistance, 
and thereby reduce the volume of current in 
tlic core to a very small value. 

A certain armature had been burned out 
by the power current that followed a lightning 
discharge to the core. The damaged coils 
were removed, the core cleaned and scraped, 
and new coils installed. The machine ran 
without trouble for several months and then 
burned out again in the same place. Attrib- 
uting the second failure to poor repair work, 
about twice as many coils were removed as 
before and the machine again repaired. 
Before it had a chance to fail again, the 
operator had a whiff of burning insulation, 
investigated and found excessive heating, 
which was confined to the repaired area. 

An "armature man" was called in. On 
removing the repaired coils, inspection dis- 
closed that the insulation armor of some of 
them was being burned from the outside; the 
cotton insulation on the wires was in good 
condition, showing that the trouble did not 
come from within. The armature was 
stripped, the core disassembled to and in- 
cluding the burned area, several inches of 
new laminations installed, and the coils re- 
placed. On drying out and again placing the 
machine in service, no tendency to heat more 
in one place than in another was exhibited. 



The cause of the heating had been the 
burning together of the laminations; the 
effect was equivalent to having a part of the 
core made of solid iron. 


Exciters for alternators may be steam- 
driven, water-driven, or motor-driven; or 
the exciter armature may be wound upon an 
extension to the alternator shaft or may ln- 
belted to that shaft. In either of the two 
latter cases, an}- factor that affects the speed 
of the alternator will directly affect that of 
the connected exciter. With an independently 
driven exciter, low alternator speed will not 
affect the exciter speed, excepting insofar as 
the increased exciter current required may 
overload the exciter and reduce its speed for 
that reason. In any event, low alternator 
speed means more exciting current to main- 
tain normal alternator voltage. With low 
exciter speed, the exciter field current must 
be increased to maintain the exciter voltage, 
even if the alternator speed is normal. There- 
fore, with direct-connected sets, low alter- 
nator speed lowers the alternator e.m.f., not 
only because the relative motion between 
armature conductors and the magnetic lines 
is slower, but because the lower exciter speed 
produces lower exciter voltage, hence less 
exciting current and a weaker alternator field 
for the alternator armature conductors to cut. 

An operator once complained that he could 
not maintain his switchboard voltage, even 
with the exciter and alternator rheostats all 
cut out. What was really needed for the 
fluctuations of his load was automatic voltage 
regulation (which he afterward installed); 
but, to meet the immediate requirements, an 
inspector analysed the conditions and found 
most of the trouble to be due to low speed 
of the alternator and connected exciter. 
Readjustment of the engine governor to give 
normal speed at full alternator load and 
more attention given to keeping up the steam 
pressure during the load swings, which lasted 
for an appreciable time, improved the 
operation considerably. 


In the older types of armature core con- 
struction a single key, in conjunction with 
the end-plates, was relied upon to hold the 
core laminations securely. For the work to 
which motors were then assigned and for 
the comparatively moderate speeds that then 
i iled, this method was satisfactory. 
Today, however, speeds are high and motors 

are applied to carry practically every type 
of load; consequently, their mechanical fas- 
tenings are designed accordingly. Occasion- 
ally, one of the old types of motor (many of 
which are still in use) is applied to work 
which it cannot continue to perform because 
it was not designed for that service; trouble 
soon follows. 

As an example of such a case an inspector 
was called to determine the cause of sparking 
and eating out of the commutator mica of a 
large 500-volt bipolar direct-current motor 
that recently had been applied to operate a 
stone crusher. The crusher was equipped 
with a very insignificant flywheel. As the 
operator had just resoldered the armature 
leads thereby eliminating possible poor con- 
tacts, the inspector ripped off the armature 
hood and examined the leads themselves. 
Several of them were found to be broken but 
their ends still made contact; several other 
leads were about to break. After blocking 
the armature shaft and the crusher so that 
the shaft could not rotate, he applied a crow- 
bar to the core and found the core to be loose 
on the shaft. Of course, there was no alter- 
native but to reassemble the core. 

The sparking had been due to open cir- 
cuits which had been caused by the relative 
movement between the commutator and the 
core breaking the armature leads. The 
service was improved further by later increas- 
ing the size of the flywheel and by providing 
a special water-rheostat to start the outfit 
without unduly overtaxing the motor. 


Belts, as well as motors, are liable to be 
gradually overloaded without the fact Vicing 
realized that abnormal service is being called 

A small alternator from which a number of 
single-phase motors were to be supplied with 
energy for operating printing presses, cutters, 
etc., was recently installed. It was found 
impracticable to run the printery with the 
motors because they would not hold up their 
speed under the conditions normally to be 
expected. The operator and all concerned 
of course blamed the generator and motors, 
because they were the new part of the outfit. 
An inspector was called in and he diagnosed 
the trouble as a case of general belt slipping. 
The prime-mover was a low-head waterwheel 
of ample size, but the several pulleys used in 
the transmission were small and their speeds 
were low. One intermediate six-inch belt, 
which was running at but 000 feet per minute, 



was called upon to transmit 25 h.p. or more. 
When an attempt was made to start the 
largest motor while other loads were active, 
the exciter field fell to almost zero on account 
of reduced speed. The pressing of an iron 
pipe against one belt as an idler increased 
the alternator speed 400 r.p.m. : the same 
application to another belt increased the 
alternator speed another 100 r.p.m. By 
systematically tightening the belts through- 
out the line, the alternator and exciter speeds 
were brought to normal value at moderate 
loads, but still the speeds would not hold up 
under the conditions imposed by taking a 
heavy cut on the cutting machine. It was 
necessary to substitute larger belts and 
pulleys in two places. 

Excessive belt tensions are of course to be 
avoided; but, inasmuch as loss in production 
is equally objectionable, the speed of the 
prime mover and its dependent belted ma- 
chines should be periodically checked and, 
if the efficiency of transmission is shown there- 
by to be immoderately low, measures should 
be taken to eliminate the defect so far as is 
possible. This axiom applies of course irre- 
spectively of whether the machines concerned 
are electrical or mechanical. 


A conservative ' ' rule of thumb ' ' often used 
for checking the safe carrying capacity of 
leather belts is: One inch of single belt 
traveling at the rate of 1000 feet per minute 
will transmit one horse power. It is assumed 
in this statement that the belt is tight 
enough to prevent slipping. An operator 
once complained that his elevator speed was 
surging and that, under the heavier but 
permissible loads, the speed dropped suffi- 
ciently to interfere with his production. As 
the armature of the driving motor had just 
been repaired and as the elevator had given 
no trouble up to the time of this repair, the 
motor was blamed for the trouble. 

An inspector could find nothing electrically 
wrong with the motor, but he did determine 
that the motor was 50 per cent overloaded 
when the elevator was loaded to its rating. 
This overload did not last long enough, how- 
ever, to do any harm. Operating conditions 
were favorable as the motor ran continuously 
and the elevator was controlled by means . <\ 
tight and loose pulleys. At the time of 
shifting to the tight pulley with a loaded 
elevator, the motor drew 70 amperes at 115 

volts, which corresponded to approximately 
10 h.p. The five-inch motor pulley turned 
at 1450 r.p.m. when under heavy load; this 
meant a belt speed of 1900 feet "per minute. 
From the rule given the single belt could 
transmit 1.9 h.p. per inch of width at this 
speed. The four-inch width, then, was good 
for 4X1.9 = 7.0 h.p. 

As a matter of fact the belt was earning 
more than this amount. Tightening it 
increased its capacity but, to make assurance 
doubly sure, a two-ply belt was substituted 
for the single one and all trouble ceased. 


Operators who are accustomed to running 
reciprocating pumps may be surprised at 
some of the characteristics of centrifugal out- 
fits. One of the differing features is the rate 
at which a centrifugal pump's output will 
increase with the speed. 

An operator purchased a motor-driven 
centrifugal pump set and installed it. 
Immediately after starting it up the motor 
sparked badly and in a few minutes was so 
hot that it smoked. (If the operator had 
known as much about the output of a cen- 
trifugal pump as he did about that of a recip- 
rocating pump he would have suspected the 
cause of the trouble at once.) As is usual in 
such cases, the motor was blamed and an 
inspector called in to locate the origin of the 
trouble. An ammeter cut into the supply 
wires showed that the motor was heavily 
overloaded and a rough measurement of the 
water delivered by the pump showed the 
amount was far in excess of that which the 
operator had specified. 

In measuring the speed, however, to check 
some readings, the speed was found to be 
greater than that ordinarily corresponding to 
about three-quarters load; this of course 
could not be the case with a perfectly normal 
accumulatively-connected compound-wound 
motor, the speed of which drops rapidly on 
increase of load. The field connections were 
checked and found to be correct, but in check- 
ing them the inspector noticed that one of the 
brush-holder insulating washers was 'broken. 
The operator explained that it had received 
a blow at the time of installing the set. It 
developed that at the same time, the whole 
brush-holder construction had been forced 
around against rotation, thereby enabling 
the armature reaction to weaken the motor 
field and thus increase the motor speed. 



By J. Murray Weed 

Transformer Engineering Department, General Electric Company 

A quantitative knowledge of the magnetic stresses in transformers is becoming more and more important 
as the size of the apparatus and system increases. In this timely article the author discusses various factors 
which have a bearing on the magnitude of the magnetic stresses at short circuit, and their bearing upon the 
size and cost of transformers. He points out that the stresses in large transformers are much greater than in 
small ones of a similar design, and therefore that an increase in the cost, for an equal degree of safety, must 
necessarily result. Also, the greater the capacity and the lower the voltage of the transformer the greater 
will be the extra cost for limiting the stress to a safe value. Therefore, in the latter case it is often more eco- 
nomical to introduce external reactance than to have it self-contained in the transformer. He concludes by 
checking his equations with equivalent ones derived by Dr. Steinmetz, and gives experimental verification of 
the formula?. This article is a continuation of the matter given under the subject "Magnetic Leakage in 
Transformers" in the December, 1912, and the January, 1913, issues of the Review. — Editor. 

The following formula for the calculation 
of mechanical stresses in a transformer was 
derived in the former article and appeared 
as equation 28. 


2.82X10- 7 «, 


lb. per sq. inch (1) 

This formula gives the stress due to the 
leakage field in any gap between coils (see 
Fig. 1) in terms of the number of turns 
effective at the particular gap, the length 
of leakage path and the current flowing. 

The term >;,. I equals ampere-turns effec- 
tive between coils where the stress is calcu- 
lated, the current being that for which it 
is desired to calculate the stress, presumably 
the short-circuit current. The length of 
the leakage path, /, is taken empirically as 
Vkh. (See Fig. 2.) 

The preceding formula is very convenient 
for calculating the stress when the limiting 
value of current is known; but when the 
current is limited by the reactance of the 
transformer, with full voltage applied, the 
following formula? are of convenience to 
the designer in showing just how the design 
may be altered to the best advantage for 
reducing or limiting the stresses. 

Confining n g in equation (1) to its maxi- 
mum value, which occurs between the 
primary and the secondary coils, we may sub- 
stitute for I the value obtained by dividing 
the voltage of the transformer by its reactance 
(neglecting the resistance component of the 
impedance) . 

The approximate formula for reactance, 
as obtained from equation (9) of the former 
article (Dec. 1912 issue of the Review) is 

■2fnHmlt)d^_ (2) 



10' I G 

where / is the frequency, and n the total 
number of turns in the transformer winding. 

The term (mlt) stands for mean length of 
turn (see Fig. 2) and is one dimension of the 
leakage area and d the other. 

is the distance between primary and second- 
ary windings plus one-third of the distance 
through the high-tension and the low-tension 
coils (including ducts) of a single group of 
coils (see Fig. 1). G is the number of such 
groups of coils. 

Equation (2), and those which follow in 
this article, apply only when the number of 
turns in all of the coil groups of the trans- 
former are equal, so that the total number 
of turns in a transformer is 

n — n s G ( 4 ) 

Attention should be called with emphasis 
to the fact that, since the maximum stress 
in the transformer is that found in the gap 
between high-voltage and low-voltage coils 
where n g is maximum and is proportional to 
(rig max.) 2 , no transformer is properly de- 
signed from the standpoint of mechanical 
stresses which does not have equal numbers 
of turns in all high-voltage — low-voltage 
groups. Unequal grouping will increase the 
reactance somewhat for a given number of 
groups, thus reducing the factor / in equation 
(1), but n £ max. will be increased much more 
than / will be reduced. Unequal grouping 
is also unfavorable from the standpoint of 
eddy-current loss. Only designs of equal 
grouping are considered in this article. 

From equation (2), we now have 

T- E - 10? l G E 

X 2fn* (mlt) d {o) 

whence, substituting this value of I in equa- 
tion (1) 

5„„„ = 7.05 X 10 6 X ^ ( ^ )2c/2 lb. per sq. in. (6) 



The only quantities appearing in equation 
(6) that can be changed in the design are 
n, (mlt) and d. These quantities appear 
also in equation (2) for the reactance. A 
study of these two equations will show how 
the stress may be limited to a required 
safe value with a minimum increase in the 
reactance. Attention must also be given, 
of course, to the effects which the changes 
in the quantities n, (mlt) and d have upon 
the other characteristics of the transformer, 
such as cost and efficiency. The designer 
will note, also, that these quantities are not 
independent of each other, as an increase in 
n for instance with constant magnetic density 
in the core will result in a decrease in the 
cross-section of the core and consequently in 

Fig. 1. A Sectional View taken through the Conductors 
of a small Shell-type Transformer 

the dimension (mlt), though the decrease in 
this dimension will not be in the same pro- 
portion. It requires one who is familiar with 
all of these relations to interpret the equa- 
tions to the best advantage for any particular 

An increase in the number of turns, with a 
corresponding reduction in (mlt) to maintain 
constant density in the core, will increase 
the reactance by a larger factor than that by 
which the stress is reduced, since (mlt) 
appears in the numerator of equation (2) 
in the first power and in the denominator 
in equation (6) in the second power. If the 
size of the core is not changed, keeping (mlt) 
constant and allowing the flux density to 
decrease as the number of turns increases, 
the factor of increase in the reactance is the 
same as the factor of decrease in the stress, 
since n appears in the second power both in 
the numerator of equation (2) and in the 
denominator of equation (6). 

Having a constant flux density in the core, 
allowing the dimension (mlt) to reduce as n 
increases, the cost will increase or decrease 
depending upon the number of turns as 
compared with the cross-section of the core, 
since there is a definite relation of these two 
quantities which gives the minimum cost. 
On the other hand, their relative values may 
be varied through a considerable range on 
either side of this most economical relation 
with but a small increase in cost. But if 
the design is varied too much, by increasing 
the number of turns with a corresponding 
reduction in the cross-section of the core, 
the cost will begin to increase rapidly. 

Where the cross-section of the core remains 
constant, permitting the flux density to 
decrease as the number of turns is in- 
creased, the cost of the transformer will 
always increase with increased number of 

Turning our attention now to the factor 
d, a decrease in the number of high-voltage — 
low-voltage coil groups will increase the A' 
and Y terms of this factor. Since these 
terms are divided by 3, however, and since 
no change is made in the Z term (distance 
between primary and secondary) , the increase 
in the value of d is much smaller than the 
decrease in G. It is seen, therefore, that the 
increase in reactance is much larger than the 
decrease in stress since, while d appears in 
the second power in the denominator of 
equation (6) and only in the first power in 
the numerator of equation (2), G is found in 
the denominator of equation (2) and not at 
all in equation (6). This change will always 
result in a reduction of cost, however, since 
the necessary length of the magnetic circuit 
is reduced by the elimination of some of the 
insulation spaces between the high-voltage 
and the low-voltage groups. 



On the other hand, if the value of J is 
increased by increasing the distance between 
the high-voltage and the low-voltage wind- 
ings, without changing the number of groups 
(increasing the Z term only), the decrease 
in the mechanical stress at short circuit is 
large as compared with the increase in 
reactance. Moreover, since this term (Z) 
enters into d at full value, this is an effective 
manner of reducing the stress. It results in 
an increase in cost, however, due to the 
necessary increase in the length of the 
magnetic circuit. This is often the only 
method by which the stress may be limited 
to the desired value without excessive re- 
actance, and without excessive loss due to 
eddy currents, which, as shown in the former 
article (Jan., 1913, issue of the Review i is 
proportional to the square of the maximum 
density of the leakage flux, or to ;//. It is 
always necessary to consider this loss in 
high-reactance transformers and it must be 
calculated as well as the mechanical stress 
and the reactance. 

Equation (6) may be somewhat simplified, 
reducing the number of constants to a single 
numerical value, by substituting the value 
of E from the fundamental voltage equation 
of the transformer, namelv 

E = 


_ \ 2 ir fn 4> 


5,„ a ,= 1400, 

<t> 2 

\mltyd 2 ( g ) 

where <p is the total flux of the transformer, 
i.e., its maximum value. 

Although not so convenient as equation 
(6) for studying the relation of the mechanical 
stress las affected by various factors of 
design' to the reactance of the transformer, 
equation (S) shows more clearly the direct 
relation of some of these factors to the 
mechanical stress. 

This equation shows again that the stre-- i - 
reduced by reduction in the flux, with (mlt) 
constant, which means an increase in the 
number of turns with constant cross-section 
re. It shows also that the stress is 
reduced by increase in (mlt), with constant 
flux, which means a constant number of 
turns and an increased cross-section of the 
The flux density of the core is reduced 
in either case, which shows very clearlv 
that high flux densities in the core' result iii 
increased mechanical sti With a given 

he core, the - pro- 

nan of the flux density. 

This conclusion is based upon the assumption 
that the change in the number of turns 
accompanying the change in flux does not 
affect (mlt) or d, which is not correct unless 
it can be effected by the elimination of an 
entire high-voltage — low-voltage group. 

The effect of increased kv-a. capacity 
upon the mechanical stresses may be seen 
from equation (S). Thus, if we start from a 
normal design for a 2.30 kv-a. transformer 
and increase all dimensions alike, keeping 
the same magnetic and current densities in 
core and windings, and assume that the 
space factor is not changed, a 4000 kv-a. 
transformer will have all the dimensions 
doubled, and the total flux <j> will be multi- 
plied by four. The two factors in the denom- 
inator, which are dimensions, are both 
doubled, while the factor in the numerator 
is multiplied by four. The stress is, therefore, 
the same as before. This conclusion is based 
upon the assumption that d was doubled, 
which means not only that the space factors 
within the high-tension and the low-tension 
groups of coils have not been reduced by the 
increased size of the transformer, but also 

low Vo/tage 




Fig. 2. 


= = 


\\ (mlt.) 

A Sectional View Parallel to the Conductors 
of a small Shell-type Transformer 

that the distance between the high-tension 
and the low-tension has been doubled. Since 
the space factors are actually reduced, the 
distance between the high-tension and the 
low-tension must be more than doubled to 
give this result. Also, if in accordance with 
common practice, the number of groups has 
been increased in the larger transformer. 



which would reduce the value of d, the 
distance between the high-tension and the 
low-tension must have been still further 
increased to give the same value of d as that 
used before. the grouping was changed. Nor- 
mally, from the standpoint of insulation, the 
distance between the high-tension and the 
low-tension would be the same for the large 
transformer as for the small one. It is 
thus seen that with otherwise normal design 
the stresses will be much larger in the large 
transformer than in the small one and that, 
in order to limit them to the same value, the 
cost of the transformer must be considerably 
increased above that of the otherwise normal 

The greater the capacity, and the lower 
the voltage, the greater will be the extra 
cost for limiting the stresses. There is a 
point where the cost will be less to supply 
an external reactance in connection with 
the transformer, for limiting the current and 
thereby the stresses in the transformer, than 
to adopt the abnormal and expensive design 
that would otherwise be required for this 

It is interesting to note that if the value 
of / in equation (4) be substituted for but 
one of the factors in equation (1), the result 
will be 

F max = 14 1— j-r. lb. per sq. inch. (9) 


This is the equation obtained by Dr. Stein- 
metz in his paper on "Mechanical Forces in 
Magnetic Fields."* 

Dr. Steinmetz's derivation for this formula 
is as follows: 

Assume that the secondary group is moved 
toward the primary group until the centers 
of the groups coincide. The turns of the 
primary coils cut the leakage flux producing 
a voltage 

The work done during this motion is 

w '= j eldt = nl<t> X 10~ 8 joules 


whence w = PL 

The energy stored in the magnetic field, 
which has been eliminated was 


W\ = 


The mechanical work done is 

u>2 = w-wi= —— =Fdg 10~ 7 joules 

where d is a distance in centimeters corres- 
ponding to the force F in grams which 
existed in the initial position, and g the 
acceleration due to gravity is 981. This 


F =it iQ7 



Substituting the value of L from the equation 

E = 2ir fLI 
we obtain 


47T fgd 

F= - )' grams per sq. cm. 

Reducing to practical units, 

per sq. inch 

F = 0.705=5-lb. 


where d is in inches. This value of F corres- 
ponds to the effective values of voltage and 
current, and is the average force. The 
maximum value will be proportional to the 
product of the maximum values of current 
and voltage, so that 

F max = 1 .41 — r lb. per sq. inch. 

This force corresponds to a single group of 
coils, the voltage E being the proportional 
part of the total voltage of the transformer 
corresponding to this group. If E is the total 
voltage of the transformer, and the groups 
of coils all have equal numbers of turns, and 
are equally spaced so that the total voltage 
will be divided equally among them, the 
force corresponding to each group, and there- 
fore to the entire transformer, is 


/<"„,,„■ = 1.41-^= lb. per sq. inch. 

This equation does not show the distribution 
of the stresses within the .transformer, nor 
the effects of the factors of design upon the 

Tests have been made to confirm the 
calculations made for the total force developed 
in a transformer. The formula? for these 
calculations, which have been given in the 
former article, are 

_ 7 wg 2 J 2 MQ_ 

= 2.82X10" 





F,;,, = 1.41 = lir 

*A.T.E.E. Proceedings, December, 1910. 

. ng"P\ uilt) 

pounds (11) 



A description of the tests referred to will 
now be given. Two sets of low-tension coils 
from 25-cycle, self-cooled. 200 kv-a. trans- 
formers of 33,000 /3300-volt rating were 
assembled loosely on the same core, with 


Fig. 3. A Photograph Showing the Scheme Employed for 

Experimentally Testing the Correctness of the 

Mechanical Stress Calculations 

pressboard collars between the coils. With 
the core on its side and the coils in a hori- 
zontal position, channel irons were placed 
through the openings supporting a platform 
upon which weights were placed. The eight 
coils were connected, four as primaries and 
four as short-circuited secondaries, and, with 
different weights upon the platform, the 
voltage and current were increased until 
the weights were lifted (see Fig. 3). As the 
current increased a point was reached where 
vibration began. The vibration became more 
and more violent until the weights were 
actually lifted. The current which caused 
vibration to begin was that giving a maximum 
force equal to the weight lifted (equation 10). 
while the current which lifted the weights 
was that corresponding to an average force 
equal to the weight (equation 11). Up to a 
point where primary and secondary coils 
began to separate, lifting the weights, the 
current and applied voltage were directly 
proportional to each other. If the voltage was 
increased beyond this point, the distance 
between the primary and the secondary coils 
was increased, giving increasing reactance, 
while the current increased very little. The 
tit to which the weights were lifted 

depended upon the separation between the 
primarv and the secondary which was required 
to give a reactance voltage corresponding to 
the applied voltage. The data for the calcu- 
lation of the forces by equation -(11) are. in 
this case, 

(mlt) =9.65X12 = 116 in. 

Z=8M in- 

«g = 41 Xnumber of coils effective. 

The results obtained from these tests are 
given in Tables I and II. 


Coil Arrange- 




Per Cent 







* 400 




























1 765 



6 1 5 
















































Coil Arrange- 











21 12 

i nil) 









Per Cent 







The tabulated weights in the "lifted" 
column include the weights of the coils, the 
channel irons, and the platform. The read- 
ings were necessarily rough, since they had 
to be taken hastily in order to prevent the 
weights from being shaken off the platform 
and the coils from getting too hot. The 
weights indicated by the asterisks(*) were 
not actually lifted, the tests having to be 
stopped too soon for the reason stated. 


By H. I. Holleman 
Construction Engineering Department, General Electric Company 

Refrigeration as applied to the preservation of perishable products has long been practiced, but the 
attention which has been given the art as a means of increasing human efficiency and comfort has not been 
commensurate with its possibilities. The following article briefly reviews past and present work, then pre- 
dicts the future developments of refrigeration as applied to bettering living conditions, and finally describes 
the benefits which will be derived therefrom by the electrical industry. — Editor. 

The addition of heat to various substances 
under various conditions for various purposes 
that tend toward the comfort and general 
advancement of the human race, is a practice 
which dates from pre-historic times. On the 
other hand, the extraction of heat, which, 
taken in its broader sense is of but slightly 
less importance, has been left to the present 
generation to bring into its fullest development. 

The addition of heat had an early origin 
probably because of the fact that it was 
practically necessary, that the process was 
easy, and that the needed materials were 
readily obtained; whereas, the delay in the use 
of heat extracting mediums can be attributed 
to the fact that practically the reverse con- 
ditions existed. Today, however, the effi- 
ciency of every individual and the preserva- 
tion of the perishable products of his labors 
are of such vast importance that heat extrac- 
tion or refrigeration has become a necessity. 
Its practicability has been secured through the 
advancement of science, which has put 
within our hands the necessary mediums 
and machines for its successful accomplish- 

The great drawback now is the cost. This, 
however, is of small importance compared to 
the benefit derived; and it merely remains to 
give the public at large the correct viewpoint, 
before refrigeration will become practically 
as common as heating. The value of refrig- 
eration to humanity is impossible of expres- 
sion in terms of dollars and cents. For 
instance, what is the value of a ticket for a 
theater which is too warm for comfort as 
compared to one for the same theater when 
refrigerated to a comfortable temperature.' 
Or what is the relative efficiency of men work- 
ing in rooms at 70 deg. F. and 90 deg. 
respectively? Or how much more fit for a 
day's work is a man who has slept in a room 
at 70 deg. F than a man who slept in a room at 
90 deg? The value of refrigeration as a pre- 
servative of food products is well recognized 
and yet there is no way of estimating the vast 
saving accomplished each year in this manner. 

There are a great many ways of refrigerat- 
ing; in fact, so many that this article could 
not cover even a brief description of them. 
However, we can classify refrigeration under 
three general heads as follows : 

1 . By natural ice. 

2. By refrigerating mixtures. 

3. By mechanical refrigeration. 

Refrigeration by natural ice is of great 
importance in the colder climates. The 
unreliability of the supply and the high cost 
of transporting the ice to the wanner climates 
make this form of refrigeration of small 
importance, however, when the subject is 
considered as a world-wide proposition. 
According to some of the most eminent 
authorities, the application of this form of 
refrigeration (even in the colder climates) is 
rapidly on the decline and it is predicted that 
the near future will see mechanical refrigera- 
tion replace that by natural ice in the majority 
of its uses. 

The use of refrigerating mixtures is of still 
less importance than the use of natural ice. 
In fact, at its present development, it is of 
practically no commercial value. 

This then brings us to the last general class, 
which is the one most widely used today and 
the one upon which we must rely for future 
development. Under this head, we have two 
subheads which include all mechanical refrig- 
erating systems that are of real commercial 
importance, viz.: 

1. The compression system. 

2. The absorption system. 

The compression system is the one in which 
a gas is compressed and cooled under compres- 
sion (usually to a liquid) and is then expanded; 
and, in expanding, absorbs heat from the 
material to be refrigerated either directly or 
indirectly. The expanded gases are then 
drawn into the compressor again and thus the 
cycle is completed. 

The absorption system is the one in which 
the expanded gas is absorbed by a liquid whose 
evaporating point is much higher than that of 



the refrigerant. The mixture is then heated, 
driving off the refrigerant as a gas at a high 
pressure. This gas is then cooled under com- 
pression (usually to a liquid) and expanded as 
in the compression system, the expanded gas 
being absorbed in the simpler systems by 
another tank of absorbing fluid which in turn 
is heated. 

Each of these systems requires power in 
some form in order to operate. The com- 
pression system, which bids fair to become the 
most popular, requires more; and herein lies 
one of the greatest opportunities for the future 
expansion of the electrical industry. The 
uses of refrigeration are many, and the total 
power load required when developed is 
enormous. The applications of refrigeration 
can be classified under two of the general 
heads mentioned before, viz. : 

1. Refrigeration for the preservation of 
perishable products. 

2. Refrigeration for the comfort and 
increased efficiency of humanity. 

There has been so much said and written 
about the former, that the importance of the 
latter has been neglected. It is, therefore, to 
this latter class that it is wished to call 
attention. The one great function in this 
field is the cooling of workshops, offices, 
sleeping rooms, public halls, etc. When we 
consider that such a tremendous portion of 
the habitable surface of the earth (the tropical 
and semi-tropical countries) is more in need of 
refrigeration than heat for buildings, we are 
astounded to see what little progress has been 
made in this direction. There is, of course, a 
greater cost to refrigeration but it is not as 
great as we would at first imagine. 

If we compare heating with refrigerating 
it will be seen at a glance that, in comparison 
with direct heating from coal or wood, the 
cost of refrigeration will be excessive, since 
the B.t.u. in generated power is seldom 
greater than 12 per cent of the B.t.u. in coal. 
However, if we compare refrigeration on the 
same basis as heating (that is, refrigerate and 
heat with same form of power, say for instance 
electricity) it will be found that much less 
1 ii iwer is required to refrigerate under average 
summer conditions than to heat under average 
winter conditions. 

Take for example the heating of a room in 
winter as compared to the cooling of it to a 
ifortable temperature in summer. 

i room 12 ft. by 15 ft. l.\ 10 ft., 
with an a -adiation of 0.2 B.t.u. per 

hour per degree difference per sq. ft. surface, 

and with one change of air per hour at 1 
B.t.u. per 50 cu. ft. of air degree rise in 

Radiating surface of the room = 450 sq. ft. 

Cubical contents of the room= 1800 cu. ft. 

Temperature at which the room is to be 
kept = 70 deg. F. 

Take the average temperature of atmos- 
phere in the winter as 30 deg. F. 

Taking the average temperature during the 
summer months as 90 deg. F. 

Assume the cost of electric current as 2 
cents per kw-hr. 

The heat units per hour necessary for 
maintaining the room at 70 deg. F. during 
the winter amount to 



On the other hand, the heat units per hour 
necessary to be absorbed from the room in 
order to maintain it at a temperature of 70 
deg. F. during the summer amount to only 

'+ 1 --!!'-V c J0-70) = 1S70 B.t.u. 


To reduce these figures to comparative 
costs, we find that the 3744 B.t.u. per hour 
for heating corresponds to 1.1 kw-hr., which 
at 2 cents per kw-hr. equals 2.2 cents per hour 
or 52.8 cents per day of 24 hours. 

According to the best authorities, we find 
that a small ice plant will produce a ton of ice 
for $1.50 per ton (including all overhead 
charges etc.), when buying power at 2 cents 
per kw-hr. A ton of ice represents a cooling 
effect of 364,000 B.t.u. From this we see 

that one ton of ice will cool 


1 51 1 

= 141 such 

rooms, or each room will cost : rjr = 1.06 cents 

per hour or 15.45 cents per 24 hours for 
refrigeration as contrasted against 52.8 cents 
for heating electrically. 

Applying the above to hotels, as being the 
easiest way to estimate the value of refrigera- 
tion, we feel safe in saying that that portion of 
the traveling public which would refuse to pay 
15 cents a day for a refrigerated room is quite 
small. Also, when we take into consideration 
the fact that the size of the plant necessary 
for hotel refrigeration would be large and 
would produce refrigeration more cheaply 
than the plant under consideration, and that 
the rooms under most conditions would only 
need refrigeration for 10 hours per day, and 
that hotels can be built with a much smaller 


radiation factor than those of today, it is easy 
to see that the actual cost per room for refrig- 
erating a hotel will be very low in comparis m 
to the comfort derived. The same is true- 
when applied to theaters, churches, sleeping 
rooms, workshops, office buildings, etc., but 
each needs to be approached from a different 
angle in order to properly realize the value 
in each case. 

It has been predicted, by those in best 
position to know, that refrigeration will 
some day be to the South what heating is now 
to the North. The uses of refrigeration in 
tropical countries are almost innumerable; 
and even in the southern part of this country 
we find that the physical and the mental 
well-being of a large portion of the popula- 
tion suffer from continued excessive heat. 
In that section of this country, the refrig- 
eration of sleeping rooms, hospitals, schools, 
and public halls would be of incalculable 

It takes but a glance at the situation to 
realize that the possibilities for such use of 
electric power are vast; and it is to be hoped 
that the large power companies will soon be 
brought to realize' this condition and wage an 

active campaign for the double purpose of 
profit and education. 

It is only within the last few years that the 
central-station management has begun to 
realize the possibilities of a marked increase 
in their load factors resulting from the use of 
electric power in ice plants. 

Since that time, electrically-driven ice 
plants have sprung up all over the country, 
both as isolated plants purchasing power from 
a central station and as those in connection 
with and a part of a central station. When we 
analyze a refrigeration load, such as one of 
those previously enumerated, we find that it is 
just as beneficial for the central station as for 
the ice plant and is, moreover, much greater 
in its possibilities. It is easier to show profit 
in applying refrigeration to the preservation 
of a product than in applying it for increasing 
the efficiency of the producer; therefore, the 
producer has suffered. However, judging 
from the reports published in the engineering 
magazines of the installation of refrigerating 
plants in some of the new theaters, hotels, 
offices, buildings, etc., we believe the time has 
come when the producer's welfare will be 
valued as highly as his product. 


By G. H. Stickney 

Edison Lamp Works, Harrison", N. J. 

The lighting of factories and offices has received a great deal of attention during the past few year-;, and 
we have published several articles in the Review on this subject. Xot so long ago the matter was given 
little attention; but it did not require much time for the illuminating engineer to prove conclusively to the 
manufacturer that it was poor economy to keep down the light bills at the expense of the workman's 
efficiency and the quality of his product. The campaign inaugurated in the cause of safety has focused 
further attention on the subject of correct lighting; and in this article, which was presented at the Safety and 
Sanitation Conference held under the auspices of the American Museum of Safety, it is shown that it is not 
sufficient to provide merely enough light, but that the position of the light sources and the proper diffusion 
of the light are important factors in securing the best illumination, with the least fatigue to the work- 
men's eyes. — Editor. 

Believing as I do, that good lighting in 
factories is one of the most effective agents 
in promoting industrial safety, I especially 
appreciate the privilege of being designated 
by the Illuminating Engineering Societ y to 
address you on the subject. 

While a very conspicuous advance in light- 
ing methods has been made by progressive- 
manufacturers, notably in the iron and steel 
industry, there are still a large number of 
manufacturers who seem to regard the light- 
ing as an expense to be reduced to the lowest 
possible minimum. 

The increased appreciation of daylighl is 
indicated by the modern type of building 
construction; in which the light-finished, 
high studded workroom, with large window 

areas, often equipped with diffusing glass, 
and sometimes supplemented with saw- 
tooth roofs, permits the fullest possible 
utilization of natural light. 

It is in the artificial lighting, however, 
that the greatest progress has been made. 
The wonderful developments in high .effi- 
ciency units have greatly enlarged the 
bilities of factory lighting during the 
hours of diminishing daylight and darkness, 
or in places where daylight can not penetrate; 
so that now a proper lighting installation is 
not only an important safeguard, but an 
actual economy. Manufacturers who are 
today securing poor illumination with older 
form of illuminants, can by a revision of 
their lighting equipment, procure a good 



illumination, not only without much addi- 
tional cost, but in many cases with an actual 
reduction in the operating cost. 

The Association of Iron and Steel Elec- 
trical Engineers, in 1910, turned their atten- 
tion to good lighting as a means of accident 
prevention. They found, as a result of 
probably the most extensive investigation 
of the subject that has as yet been made, 
that a higher standard of illumination was 
demanded for efficient manufacturing than 
simply for accident prevention. Their report 
of progress, presented at the convention in 
September, 1913, showed from actual figures 
that in the last two years the amount of 
illumination furnished in iron and steel mills 
has increased 35 per cent; and in this con- 
nection, it is interesting to note that this 35 
per cent increase was accompanied by a 
reduction of five per cent in the power con- 

The economic value of good illumination, 
aside from accident prevention, is evident 
when we consider the greater facility with 
which an employee can work under good 
illumination, and the greater accuracy with 
which gauges can be read and tools set. 

One large manufacturer, on investigating 
his lighting conditions, found certain depart- 
ments in which, during the winter months, 
the operatives were practically idle for about 
an hour a day solely on account of darkness. 

Good artificial illumination can be fur- 
nished in such a factory for eight hours a 
day at a cost equivalent to about five minutes 
of the time of the workmen benefited. This 
illustrates the extravagance of poor lighting. 
If time permitted, one could readily demon- 
strate that, for a great variety of conditions, 
good illumination reduces the manufacturing 
costs by increasing production, raising the 
quality of workmanship and reducing the 
number of defective parts and "seconds." 

The question of safety as influenced by 
illumination presents two phases: First, 
the prevention of accidents; and second, the 
preservation of eyesight. While these two 
phases are often closely related, there are 
many conditions in which they are entirely 
independent of each other. The phase of 
accident prevention is illustrated in the case 
of the foundry or other shop where cranes 
or other powerful machinery are in opera- 

The liability of crane and elevator accidents 
is very much reduced with proper lighting. 

In the foundries and yards of a plant, it is 
practically impossible, even with safety 

committee inspection, to eliminate irregu- 
larities under foot. If not illuminated these 
may readily cause falls, with resulting 
injuries; and in foundries where molten metal 
is carried and hot metal abounds, they may 
often cause serious burns. 

Even though guarded to the fullest extent, 
powerful machinery — in which materials are 
machined and fashioned into articles of com- 
merce, and in which the arms and limbs are 
as readily crushed — presents a menace unless 
the operatives are given an opportunity to 
see and thus avoid the danger points. 

Xow let us consider the preservation of 
the eyesight. Although the blind are trained 
to do remarkable work in certain lines, there 
is practically no manufacturing operation in 
which a blind person is not at a disadvantage, 
while there are many which cannot be carried 
on without accurate visual inspection. Some 
of these operations produce considerable 
strain even under good illumination, and to 
require their performance under poor illumina- 
tion is certain to result in more or less rapid 
impairment of vision. While economy should 
in all cases require the best lighting practice, 
humanity demands it. 

In view of the preceding discussion, one 
might very properly ask : ' ' What is good 
illumination?" Judging from some of the 
attempts that have been made to solve light- 
ing problems, the conclusion might be drawn 
that simply a higher intensity of light is the 
answer. Undoubtedly a higher intensity of 
illumination is needed in most workrooms, 
but there are other features of equal and 
sometimes greater importance. The mini- 
mum intensity acceptable generally depends 
upon the reflecting power of the surfaces to be 
seen, the fineness of the detail to be observed, 
the time of observation and the closeness of 
application. Unless glare be introduced, a 
higher intensity of light is rarely objection- 
able, except from the standpoint of cost. 

Owing to the remarkable adaptability of 
our eyes, we are able to get along satisfac- 
torily with very much lower intensities of 
artificial light than are usual with natural 
light. The gain secured by the increase of 
intensity is not proportional to the intensity, 
and there is a point beyond which the gain 
would not warrant the additional cost. 
However, the standard 'of artificial lighting 
intensities is being raised, on account of the 
lessening cost of light, increasing cost of 
labor and overhead charges, and especially 
the increasing appreciation of the value of 



Perhaps the best way to consider the oth< r 
feature of good illumination will be to point 
out some of the most common shortcomings 
found in factory lighting. 

From my own observations, the most 
common defect is excessive glare and absence 
of diffusion. Glare is usually caused by 
bright lights in the field of vision. This 
may emanate directly from the light source 
or may be reflected to the eye by a glossy 
surface; it can also be caused wherever 
excessive contrast of intensity appears in 
adjacent fields of vision. The dazzling effect 
is not only unpleasant, but interferes with 
seeing. Under continued exposure, eye strain 
and even permanent injury to the eye may 

I have seen lights intended to illuminate 
stairways so arranged that, on descending, 
one could hardly see where to step on account 
of the glare. Such conditions are conducive 
to bad falls, whereas if the eyes were properly 
shielded from the glare, a lower intensity 
would have been ample. 

The unshielded light hung over a machine 
is a common source of eye fatigue. The 
glare may not be very evident at first glance, 
but when the workman's eyes have been 
subjected to such light for a long time, dis- 
comfort and inability to see result. 

The workman frequently complains of 
insufficient light when in reality the intensity 
may be higher than is required for the work. 
In case an attempt is made to meet the com- 
plaint by installing a larger light, the work- 
man's eyes are subjected to a still more 
severe strain. The proper correction should 
be to shield the light by means of a proper 
reflector, and as such a reflector would tend 
to direct more of the light upon the work, 
the working intensity would be increased; 
so in many cases it is possible to reduce the 
size of the lamp, or better yet, to relocate the 
lamp so as to enlarge the area illuminated. 

When a light can not be removed entirely 
from the field of vision, its brilliancy should 
be reduced by means of diffusing globe or 
reflector, so as to increase the apparent size 
of the light source and reduce the contrast 
between it and the background. This has 
the additional advantage of reducing the 
sharpness of shadows in the illumination, a 
result which is of considerable importance in 
rendering the various parts of a machine or 
other object readily discernible. 

Glare received from specular reflection of 
glazed paper, desk tops, polished metal, etc., 
often induces eye trouble, headache, and other 

indispositions; though the sufferers may not 
be aware of the cause. The remedy is to 
change the relative positions, so that the 
reflected light is kept out of the eyes as much 
as possible, and to enlarge the dimensions of 
the light source, as already mentioned. 

Another defect commonly found in in- 
dustrial lighting is improper distribution. 
This may be due to too wide a spacing of 
lighting units. Under this condition some 
parts of the room are insufficiently lighted 
while other parts may have more light than 
is necessary. 

Improper direction of light may illuminate 
the wrong side of the machine, leaving the 
important parts in shadow. If the bright 
parts are near the shaded ones whatever 
illumination may fall upon the shaded portion 
is rendered less effective by contrast. 

Unsteady or flickering illumination is 
always objectionable; both on account of 
discomfort and the inability to see. Such 
variation should always be avoided, whether 
caused by the units themselves or by the 
light passing through moving wheels, etc. 

Since the purpose of the lighting is to 
enable the operative to see, good illumination 
can not be prescribed until we have some 
knowledge of the use to which it is to be put. 
In order to plan the lighting of a factory 
properly, one should be familiar with the 
processes employed, the arrangement of the 
machinery and the work tables, as well as the 
quality of the product manufactured. Prac- 
tice has established certain methods of light- 
ing which, if properly applied, are satisfactory 
for the different processes of manufacture. 
Thus we know approximately how much 
illumination is necessary for the ordinary 
grade of work as performed on a lathe, as 
well as the direction desirable. As far as 
possible, therefore, the experience gained in 
well-lighted factories should be utilized in 
planning the lighting installation. The 
pamphlet entitled "Light; Its Use and 
Misuse" which has been issued by the 
Illuminating Engineering Society, is full of 
useful suggestions in connection with the 
lighting problem, while the pamphlet, 
"Modern Industrial Lighting" issued by the 
Commercial Section of the National Elec- 
tric Light Association, endeavors to make 
some specific application of this information. 
A similar booklet has been issued by the 
National Commercial Gas Association. Books 
and articles, manufacturers' publications, 
etc., furnish much useful data on this sub- 



Where extensive lighting problems are to 
be solved, it is advisable to retain a com- 
petent engineer with illuminating engineering 
experience. However, the following com- 
ments on various methods of factory lighting 
will give some idea of the general practice. 

The practice in factory lighting has de- 
veloped along a few fairly definite lines, 
which may be designated as localized light- 
ing, general lighting, combined general and 
localized lighting and localized general or 
group lighting. 

Localized lighting originated with the low 
power portable or semi-portable lighting 
units. These were under the control of the 
individual workman, to be placed or shifted 
wherever he desired. Such lamps were com- 
monly used without reflectors and produced 
small patches of uneven illumination, as well 
as more or less glare. In many cases lighting 
with these lamps is now being supplanted by 
other methods, on account of the following 
disadvantages. Lamp breakage is likely to 
be high, and the expense for installing, 
energy supply and maintenance excessive, 
depending upon the conditions and arrange- 
ment of work. Moreover, the attention of 
the workman is called to the lighting and 
much time is often lost from his regular work 
in adjusting the lamp. There are, however, 
certain operations which require light inside 
of a small cylinder or other enclosed space; 
or where very high intensities are required 
over small areas, and for these no other 
method is as practicable as localized lighting. 
For such conditions, the lamp should be 
equipped with a reflector to shield the work- 
man's eyes and reflect the light in useful 
directions. Localized lighting should also 
be used in connection with general lighting, 
as referred to later. 

"General lighting" came into common 
practice with high power lamps. Since with 
these units economy makes a wide spacing 
necessary, the best method of applying them 
is to equip them with diffusing globes and 
reflectors, so arranged as to distribute the 
illumination as evenly as possible. Lamps 
are hung high, in proportion to their power 
and the intensity required, and equally 
spaced throughout the room. The ideal 
sought is equal intensity over the entire area. 
General lighting is provided in three principal 
ways, which are known as direct, indirect and 
semi-indirect lighting. With direct lighting, 
the larger part of the light is distributed 
directly from the lighting unit to the surfaces 
to be lighted. With indirect lighting, the 

light source is concealed and the light thrown 
upon the ceiling or wall and thence redistrib- 
uted for use. With the semi-indirect light- 
ing, the light source is shaded by a trans- 
lucent reflector and the larger part of the 
light thrown upon the ceiling or walls for 
redistribution. Direct lighting, depending 
upon the equipment, may have excessive 
brilliancy or any degree of diffusion. It is 
used to a much larger extent in factory light- 
ing because factory ceilings are seldom good 
reflectors. Direct lighting units are less 
affected by dust accumulations. The in- 
direct and semi-indirect give excellent diffu- 
sion, and are often applied with good effect 
in offices and drafting rooms when light 
ceilings are available. 

"Combined general" and "localized light- 
ing" is often desirable. With this, a low 
general illumination is supplied by large 
units and more intense localized illumination 
at particular points by low power units. The 
localized lighting may be supplied con- 
tinuously or temporarily as needed. For 
example, in lighting automatic machinery, a 
moderate illumination may be sufficient at 
all times except when a machine is being 
inspected, set up or adjusted, when a localized 
light may be needed for the particular 

"Localized general" or "group lighting" 
is a recent practice which has sprung up since 
a range of intermediate sizes of lighting units 
has become available. This practice differs 
from general lighting in that, instead of 
striving for even intensity throughout the 
room, lamps are arranged to give higher 
intensities and correct direction of light at 
the machines or tables and a lower intensity 
at intermediate points. It differs from 
localized lighting in being planned so as to 
give some illumination, sufficient for the 
needs, in all parts of the room. It is, there- 
fore, an intermediate practice between the 
extremes of localized and general lighting. 
Its application is extending very rapidly, 
since it meets effectively and economically 
factory requirements for a large portion of 
the ordinary processes and buildings. 

Each of these various methods of lighting 
has some field in which it is to be preferred 
to any of the others. The selection depends 
upon the character and construction of the 
building, the process of manufacture, the 
source of energy available and various local 

That the progress in good factory lighting 
will be even more rapid in the future seems 



unquestionable. The interest of the public 
has been indicated by the recent labor 
legislation passed in New York State; 
and the broad basis on which this is 
being undertaken is indicated by the 
fact that the Museum of Safety and 
the Illuminating Engineering Society were 
consulted with regard to those portions 

of the law which had to do with factory 

While good factory lighting is likely to be 
made compulsory by law, it is hoped that the 
manufacturers will be sufficiently . awake 
to their own interest to take any necessary 
steps of their own initiative rather than 
through compulsion. 


Standardization Rules 

A new edition of the A.I.E.E. Standardiza- 
tion Rules bearing the date of Dec. 1, 1914, 
is now in effect and supersedes the 1914 
edition. Many radical changes have been 
made. Copies may be obtained from the 
office of the Secretary of the A.I.E.E., 33 
West 39th St., New York City. 

Institute Meeting in New York, Dec. 11, 1914 

The 302d meeting of the American In- 
stitute of Electrical Engineers was held at 
the Engineering Societies Building, 33 West 
39th St., New York, on Friday, December 
11th. Two papers were presented at the 
meeting as follows: Insulator Depreciation 
and Effect on Operation, by Mr. A. 0. Austin 
and Effect of Altitude on the Spark-Over Volt- 
ages of Bushings, Leads and Insulators, by 
Mr. F. W. Peek, Jr. 

These two papers appear in the December 
issue of the Proceedings of the Institute. 


On December 2d, Prof. Elihu Thomson 
addressed a meeting of about 370 members 
on Wireless Telegraphy. 

The lecture was illustrated with numerous 
lantern diagrams. Prof. Thomson first spoke 
of very early experiments by himself and 
Prof. Houston, which were conducted much 
before those of Hertz, and which showed 
definitely the propagation of ether disturb- 
ances to distances very great in proportion 
to the dimensions of the apparatus employed. 
He then showed by means of a series of well 
chosen diagrams the close relation of wireless 
to metallically-directed transmission, and 
pointed out the difference between the con- 
ditions obtaining in a Hertzian oscillator and 
a wireless transmission. It was shown how 
one-half of the figure which represents the 
ether disturbance, in the case of the Hertz 
experiments, is absent in wireless trans- 
mission, being suppressed by the conducting 
surface of the earth. The importance of the 

conducting surface, principally the salt water 
surface of the earth, was carefully brought 
out, and the effect of dry earth masses in 
obstructing the waves was described. It was 
also shown how interference waves may occur 
when alternative paths of different lengths 
are present. 

The manner in which the wireless waves 
follow the earth's surface was illustrated, 
and Prof. Thomson explained his theory of 
why this should be as it is. It is to the effect 
that the surface electric currents which are 
necessarily positioned in the water surface 
of the earth, compel the electrostatic and 
electromagnetic waves with which they are 
untied to follow the earth's curvature. The 
losses due to corona were mentioned and an 
explanation of daylight wireless transmission 
losses was proposed, which was to the effect 
that the liberation and re-absorption of irons 
produced by ultra-violet ionization caused a 
frittering away of the energy of the waves. 

On December 14th, Mr. Howard W. 
DuBois, Consulting Mining Engineer, spoke 
in Burdett Hall to a large audience. The 
subject was Alaska, Our Land of Midnight 
Sun. The speaker outlined some of the large 
hvdro-electric projects in connection with 
mining operations, spoke of the Govern- 
ment's new railroad policy, and made ex- 
tended reference to Alaska's agricultural 
possibilities. The Alaskan coal deposits and 
the large scale mining operations in con- 
nection with low grade gold ores and the very 
high grade of copper ore in the Copper River 
district were described. The lecture 'was 
illustrated by 100 very beautifully colored 
lantern slides taken from photographs made 
by the speaker when in Alaska. 

On January 6, 1915, a paper entitled, 
Modern Views of Electricity will be read by 
Prof. D. F. Comstock of the Massachusetts 
Institute of Technology. 

On February 3. 1915, Major J. A. Shipton, 
United States Army, addresses the Section on a 
subject which will be announced in due course. 




At the November 19th meeting of the 
Pittsfield Section of the A.I.E.E. Mr. W. L. 
R. Emmet read a paper, illustrated by lan- 
tern slides, on The Mercury Vapor Turbine. 
The main outlines of the paper have been 
covered by the author in the General 
Electric Review of Januarv and Februarv, 

Prof. W. S. Franklin, of Lehigh University, 
will lecture to the Section on January 7th, 
his subject being Electric Waves. 

The Section each year conducts for its 
members, classes in advanced theory, the 
subjects this year being, Electro-chemistry 
and Electric Waves. 


The Ninth Season of the Schenectady 
Section of the A.I.E.E. was opened by an 
introductory address by Mr. F. C. Pratt on 
October 6, 1914. This was followed by an 
illustrated lecture by Mr. J. B. Taylor, 
entitled, The Color of Light. 

On October 20th a lecture was given by 
Dr. E. J. Berg, on Differential Equations used 
in the Study of Transient Phenomena. 

On November 17th a large audience was 
addressed by Dr. E. K. Mees, Head of the 
Research Laboratory of the Eastman Kodak 
Company, on the subject of Methods of Photo- 
graphic Investigation. 

On December 1st and 2d, Mr. J. B. Taylor 
addressed the Section on The Choralcelo and 
Other Electrical Musical Instruments. 

Applications of electricity to the musical 
field were considered briefly under three 
general heads. 

The use of electric motors, more as forms 
of mechanical energy, in which application 
the "blowing" of pipe organs is the most 
extensive. Automatic pianos or orchestrions 
make use of electric motors. Large solenoids 
have been used to strike the bells forming 
chimes in church towers. 

In the second group electricity is used for 
control. Here again the pipe organ is the 
typical example; contacts are made on 

pressing the keys, or on actuating the other 
devices which energize the magnets by electro- 
pneumatic valves controlling the admission 
of air to the pipes. This electric control, 
as distinguished from simple mechanical 
connections or tubular pneumatic action, 
gives quicker response, greater freedom of 
arrangement of the key-board and instrument 
proper, and affords the player a variety of 
effects and greater ease of handling. 

In the third application, of which the 
choralcelo is an example, the musical tones 
themselves are produced more directly by 
the electric currents. The telharmonium 
was referred to and described briefly. In 
this instrument a "musical central station," 
consisting of 150 or more alternating current 
generators of different frequencies produce 
music at points more or less remote from the 
center of control through the medium of 
telephone receivers and wound re-enforcing 

In the choralcelo the musical tones are 
produced on steel strings like those in a 
piano. The strings are made to vibrate 
continuously by an electromagnet placed a 
few millimeters away and supplied with 
a pulsating current of the same frequency as 
the natural vibration period of the string. 
Similarly flat bars of wood or metal, of 
proper length and weight to correspond 
to the musical scale, are vibrated contin- 
uously by the application of electro-magnets. 
The variety of tonal effects available by 
various combinations of strings and bars as 
well as further variety from applying har- 
monic frequencies was demonstrated. 

The December loth meeting was devoted to 
the subject of Abnormal Luminous Manifesta- 
tions. The speakers and their subjects were: 
"Lightning," by Prof. E. E. F. Creighton. 
and "Phosphorescence and Fluorescence," 
by Mr. W. S. Andrews. Each paper was 
accompanied bv experimental demonstrations. 

On January^, 1915, Mr. S. H. Blake will 
read a paper on Electric Illumination. Mr. 
Halvorson and others will collaborate and 
there will be experimental demonstrations 
of an especially interesting nature. 




Many mathematical formulas relating to various 
operations of electricity pertaining to transients 
indicate that the transient period never ends — as 
oscillatory current never ceases to oscillate, the 
current resulting from suddenly applying a con- 
stant voltage to a circuit with self-induction never 
stops increasing in strength, etc., at least not within 
finite time. Such equations involve exponential 
functions of e related to time, all of which lead to 
infinite time as essential to a steady electrical state. 

We are told in the textbook that after a certain 
time the oscillatory current has "practically dis- 
appeared," that in a fraction of a second, or within 
a few seconds, the difference between the rising 
current in the circuit with constant impressed volt- 
age and its value at infinite time is "negligible," 
etc. But they never imply that the theory of 
operations is defective in the slightest degree, or at 
least not in respect to the infinity of the time- 
element when the steady condition is attained. 

The question once arose in the mind of the 
writer: Do the formulas correctly express the facts 
as to time, or do conditions exist that have not been 
taken account of, which, if embodied in the formulas, 
would show that a steady condition will be attained 
in finite time? 

The possible influence of the increase in resist- 
ance due to the heating effect of the current, as an 
agent for bringing about a steady current flow in a 
finite time, was naturally thought of. Reasoning 
directly applied to the simple case of constant volt- 
age applied to a circuit having only resistance and 
self-induction, and indirectly by analogy derived 
from other domains of physics involving the effect 
of heat, at first appeared to indicate a steady cur- 
rent in finite time; and likewise in the still simpler 
case of a circuit with only resistance, in which latter 
case, however, although the current instantly 
arrives at maximum value, and thus at zero time 
instead of infinite time, it is obvious that the heat- 
ing effect of the current will at once begin to in- 
crease the resistance and decrease the current 
strength. But a little further consideration of the 
problem determined an opposite conclusion in both 
the above cases. As a matter of reasoning, the 
reader must be left to consider the subject, if he 
so chooses, in his own way. 

An effort was made analytically to test the ques- 
tion from the heat standpoint for the simplest case, 
that of current in a circuit of simple resistance and 
constant voltage: 

We then have 
ir = E, (1) 

and the well-known empirical formula r = r (1 + 
ah-\-(ih 2 ), expressing the relation between resistance 
and temperature of a conductor. To condense, we 
will omit the last term, and write 

r = r„ (1 + ah). (2) 

In fact, it would make no difference in the final 
result, so far as determining whether the current 
becomes constant in finite time, if we wrote r—$ h. 
The rate of heat generation in the circuit is Ei; 
if the temperature of the surrounding medium is 
hi, which we assume constant with no detriment to 
the accuracy of the particular problem in hand, the 

rate of heat dissipation will be expressed, with no 
inaccuracy for our purpose, by -q (h — hi); whence 
the rate of heat accumulation will be S [Ei — 
r)(h — hi)] and we have for the equation representing 
rate of temperature rise: 

~=\i[Ei- n {h-h)}. 

From (1) and (2) 
ir (1+ah) =E, 


a r i 
Differentiating (4) 

di + a h di + a i dh = 0, 
and substituting from (5) in (6) 

dh= -- 

aro i' 






Substituting (5) and (7) in (3), and reducing, we 
finally have 


[X«ijE-XSij r Q (l + a hi) i- a\S r Ei l \i 
which is of the form 

-dt, (8) 

(a+bi + ci 2 ) i 

= dt. 

p M = C'dt, 

J p(a+bi + ci 2 ) i J o 


_£ ; : 

2a° e a+bi+ci 


2ci+b-i/- ( 

2ci + b + \/ 
which q=iac 

V - Eh ( '■ 

i'\p 2a\V 

/Ji> Jo 


b 2 is < 0, and p = 

r„ (1 + a hi) 

represents the current value at zero time, when the 
conductor will be at the temperature hi of the sur- 
rounding medium. 

From the last equation, the value of the current i 
will include an exponential function of the logarith- 
mic base c in respect to time. Therefore the heating 
effect of the current upon the resistance of the circuit 
will not cause the diminishing current to arrive at a 
steady value in finite time, and obviously the same 
may be said in respect to a rising current when 
self-induction is present. 

The limiting or steady value of i in infinite 
time is 




which in the final numerical result will have a plus 

Chas. L. Clarke. 




The purpose of this department of the Review is two-fold. 

First, it enables all subscribers to avail themselves of the consulting service of a highly specialized 
corps of engineering experts, or of such other authority as the problem may require. This service provides 
for answers by mail with as little delay as possible of such questions as come within the scope of the Review. 

Second, it publishes for the benefit of all Review readers questions and answers of general interest 
and of educational value. When the original question deals with only one phase of an interesting subject, 
the editor may feel warranted in discussing allied questions so as to provide a more complete treatment 
of the whole subject. 

To avoid the possibility of an incorrect or incomplete answer, the querist should be particularly careful to 
include sufficient data to permit of an intelligent understanding of the situation. Address letters of inquiry to 
the Editor, Question and Answer Section, General Electric Review, Schenectady, N. Y. 


(125) Fig. 1 illustrates a single-phase three-wire 
distribution system using two step-down trans- 
formers. What are the requirements of the 
National Electrical Code with regard to ground- 
ing the neutrals b and e? 

Assuming that lines a, b and c are of the primary 
or high-potential side of the system, the grounding 
of b as shown in Fig. 1 is not specified in the National 
Electrical Code. The direct grounding of the 
neutral of a high-potential system is left to the dis- 
cretion of the company operating the system. The 
protection of high-potential lines by lightning 
arresters, however, is required by the Code. 







Fig. 1 

Assuming that lines d, e and /, Fig. 1, are the 
low-potential secondary of the system, the National 
Electrical Code requires that e must be grounded 
if the voltage between e-f or e-d is less than 150 
, and may be grounded if the voltage exceeds 
150 volts. 

Private industrial lighting or power plants are 
exempt from the above rules unless the voltage of 
the primary exceeds 500. 

While the National Electrical Code is non-com- 
mittal in regard to grounding the neutral of a high- 
potential system, it insists on the use of lightning 
arresters and recommends several methods for 
grounding them. The Electrical Committee of the 
Fire Underwriters has regarded the grounding of 
electric systems as a means of reducing the risk of 
shock or injury to persons, but which at the same 
time tends to increase rather than reduce the tire 
hazard. In view of the fact that the standard 
lighting voltage is now almost universally used 
by all classes of consumers, the " fire hazard " yielded 
"life hazard" only in that field of a-c. and d-c. 

service covered by a voltage not exceeding 150, and 
beyond that voltage the grounding is optional. 
Some measure of this sort was considered necessary 
because of our every-day personal contact with 
lighting fixtures and their wires, which, as a result 
of familiarity, naturally engenders carelessness. 
Since the voltage of this type of circuit usually 
ranges from 100 to 125, the arbitrary value of 150 
volts was chosen as the high limit in order that the 
entire field of ordinary lighting circuits would be 
covered, and thus a greater assurance of personal 
protection be obtained. F.A.B. 


1 126,1 If the fiber insulation in the slots between 
the bars and the rotor iron of a squirrel cage 
induction motor becomes charred: {1 ) Will 
the motor take more power? (2) Will the speed 
be affected? The question has reference to rotor 
slot insulation only, the stator windings being 
assumed to be in perfect condition. 

The electrical characteristics of such a motor, 
so far as we have been able to determine, will be 
the same whether bar insulation is used or not. 

The present method of placing copper bars in 
the rotor slots is to use no packing or insulation 
whatever. The bars and the slots are made approxi- 
mately the same size, which necessitates that the 
bars be forced into the slots under pressure. Before 
adopting this method, machines of this type were 
carefully compared in test with others of the 
insulated rotor bar type. The test results of the 
two types were so nearly alike that the rotors could 
not be identified by them. 

The whole matter resolves itself into the question: 
"What will be the mechanical stability of the bars 
in case their insulation is burned out?" If an 
insulating packing is used and later becomes burned, 
the bars might become loose in the slots and tend to 
rattle, which action may ultimately break the joint 
between the bars and the end rings. There are, 
nevertheless, a number of motors operating in this 
manner and comparatively no difficulty has been 
experienced with them. By using no insulation, 
there is no packing that can be destroyed and conse- 
quently the bars will always remain firm in the slots. 

It is to be noted that practically all motors with 
so-called "slot armor" are grounded in one or more 
places due to the sharp edges of the iron laminations 
against the copper bars, and also that the horn 
fiber is used for mechanical packing, while its value 
as an insulator is purely incidental. 




(127) Is there any trouble likely to result from 
placing the leads of both the potential and current 
transformers for a polyphase meter in the same 

Provided the insulation used on the leads is suffi- 
cient to withstand the voltage strain, we see no 
objection to this practice, for the effect of mutual 
induction of the leads upon the registration of the 
polyphase meter is too small to be considered. 



(128) What would be the effect on the character- 
istics and the heating of a two-phase, squirrel-cage 
rotor, induction motor to run it from a supply 
consisting of T-connected, three-to-two-phase 
transformers in which a teaser tap of 92.5 per 
cent is used instead of one of the correct value, 
86.7 per cent? 

A three-to-two-phase T transformer connection, 
even when employing correctly spaced taps, 
will cause a small flow of wattless current in the 
transformers. If the voltages are not correct in 
ratio, as when the 92.5 per cent tap is used, the 
amount of this wattless current will be considerably 

The effect of such a supply on the operating 
characteristics of a motor cannot be definitely 
stated for it will depend entirely upon the motor's 
design constants. The tendency of the unbalancing, 
however, will be to cause the motor to act as a 
phase converter, drawing power from the lightly 
loaded line and distributing it to the heavily loaded 

This phase-converter action will cause additional 
heating of course. Considering an average polyphase 
motor, this 7 per cent unbalancing may cause an 
increase of 30 to 40 per cent in the temperature 
of its hottest part. 



(129) What is the best and quickest method to 
employ in measuring both the hot and cold 
resistances of a large number of transformers 
when under test? 

Specify the ranges of the instruments and the 

standard resistances required. 

The method which has afforded the most satis- 
factory results under the conditions named is that 
of direct-current potential drop. In the employ- 
ment of this method a steady reliable source of 
direct current and a rapid but accurate means of 
measuring the current and its potential drop are 

As a source of supply a storage battery will main- 
tain a steadier value of current than will the usual 
generator and be more completely satisfactory as a 
whole. If the measuring set is to be constantly in 
use, it would be better to use two sets of storage 
batteries so connected by switches that while one is 
discharging the other is charging. By means of a 
four-pole double-throw switch this operation can be 
accomplished automatically. 

For measuring the current, the most convenient 
method is undoubtedly that of a milli-voltmeter 
used in connection with the shunts calibrated 
especially for it. The location of these are 
indicated by S and MV in Fig. 1. By changing the 

position of plug P, any shunt may be placed in the 
circuit so that the milli-voltmeter will register the 
drop across it. It has been found convenient, in 
measuring the resistance of the usual run of trans- 
formers, to have these shunts calibrated so that 0.15 
amperes through the smallest shunt and 50 amperes 
through the largest shunt will produce a full-scale 

For measuring the voltage drop across the trans- 
former windings a second milli-voltmeter combined 
with multipliers provides the most convenient method 
of reading from 1 to 50 volts. The multipliers are 
represented in Fig. 1 by M, the milli-voltmeter by 
MV. The key, K, serves to complete the circuit 
through the milli-voltmeter and also to prevent the 
liability of burning out the instrument by allowing 
instantaneous trial contacts to be made. 

Fig. l 

It may then be convenient to arrange two or 
more voltmeter circuits, as shown by /, ,2 and 8 in 
Fig. 1, so wired through a three-circuit switch, 
TCS, that each in turn may be connected to the 
voltmeter. The transformer windings 7\, Ti and T 3 
may then be connected in series and the resistance 
of each be measured in such quick succession as to 
be made almost simultaneously. This practice 
will be found particularly effective in measuring 
resistances at the completion of temperature tests. 

Although a milli-voltmeter may be said to be 
automatic, and consequently furnishes quick results 
for that reason, for potential drops of one volt or 
less it becomes rather unsatisfactory. Under such 
conditions a potentiometer, though slower in making 
the measurements, should be used. Its superiority 
lies in the fact that a dry battery of standard voltage 
is bucked against the drop across the windings and, 
when a balance is obtained as shown by a galva- 
nometer, there is no current flowing in the so-called 
"drop lines" (which are the voltage measuring 
lines that are tapped to the transformer windings). 

The most suitable current to hold while measur- 
ing resistance appears to be from 10 to 15 per cent 
of the current capacity of the windings which are 
being measured; this is usually great enough to 
give a good reading on the instruments and is yet 
so low that the temperature of the transformer 
windings will not be materially affected. 



It is with the deepest sorrow that we record 
the death of Douglas S. Martin, a former 
editor of the Review, which occurred in a 
military hospital at Boulogne, France, on 
Sunday, November 22d. 

Mr. Martin was wounded by shrapnel on 
the battle field at Messines near Ypres, on 
November 1st. He was carried to a field 
ambulance by two men in his own squadron, 
and his wounds were attended to at a field 
hospital. He was then taken to the hospital 
at Boulogne, where it was at first thought he 
would recover; but after three weeks of suffer- 
ing septic poisoning de- 
veloped, and on Nov. 
22d he joined the ranks 
of those who had " served 
to the uttermost." 

Mr. Martin was a 
student of the Central 
Technical College of 
London, and after his 
graduation he entered 
the employ of the Brit- 
ish Thomson-Houston 
Company at Rugby, 
England. In 1911 he 
came to the United 
States to accept the posi- 
tion of Assistant Editor 
of the General Elec- 
tric Review, and 
succeeded to the editor- 
ship in 1912. He was 
an energetic and capa- 
ble writer on technical 
subjects, and as an editor 
his initiative and person- 
ality did much to increase 
the usefulness and pres- 
tige of the Review. 

With the object of improving his knowledge 
of practical field work, particularly in regard 
to high tension transmission systems, he 
resigned his position as editor in July, 1913, 
and went to Vancouver, B.C., from which 
point he traveled south along the Pacific 
coast, working in various capacities on a 
number of engineering projects. During this 
period, he continued his literary contributions 
to the technical press, and early in 1914 re- 
turned East, becoming a member of the 
editorial staff of the Electrical World. He 
organized the more recent statistical work of 
the paper, constituting practically a new 
department, of which he remained in charge 
until his departure for the front. Although 
only 27 years old, Mr. Martin had already 
attained a high standing in his profession. 


Upon the outbreak of the European con- 
flict, Mr. Martin, who had had considerable 
military training in the yeomanry of his 
country, immediately volunteered for active 
service, and as he was an accomplished 
horseman, a good shot, and in excellent 
physical condition owing to his activity in 
outdoor sports, his services were accepted 
promptly. He was assigned to the 16th Lan- 
cers, which was part of the first British ex- 
peditionary army, so that within a very short 
period he had exchanged the quiet of the 
editorial office for the gruelling turmoil of the 
battlefields of Flanders. 
This, in a letter written 
at this time, he charac- 
terized as the "greatest 
of good luck." 

Again, in a letter dated 
Monday, August 31, 
1914, he says: "I got 
home here last Friday 
afternoon. I signed up 
for Kitchener's Army 
on Saturday, and took 
the shilling this morn- 
ing. I expect to be de- 
tailed to some regiment 
tomorrow and, of 
course, shall be with 
them till the end of the 
war." He wrote again 
on September 20, 1914. 
"I am working myself 
awfully hard with sword, 
lance and rifle, so that 
I can get away with 
one of the early drafts. 
Tired but fit." 

Although the bones 
of the lower leg were 
broken, the knee cap injured, and the muscles 
and tendons of the leg torn, in writing from 
his bed in the Boulogne hospital he made 
light of his wounds, representing them as 
being slight. 

Douglas Martin was a young man of finest 
qualities; his engaging personality, finished 
and able conversation, and his accomplish- 
ments in vocal and instrumental music won 
for him many admiring and devoted friends. 
Talented, lovable, and loyal to the core, 
with the promise of a brilliant future, his 
untimely death is a great and irrepar- 
able loss to all who knew him well. The 
sincere sympathies of his many friends in 
America are extended to his mother and 
the other members of his family in their 

General Electric Review 


w w n r>r^c r,,^ InIIM t, „„„.„_„ Associate Editor, B. M. EOFF 

Manager, M. P. RICE Editor. JOHN R. HEW ETT ..... ' „ .,„.„.- 

Assistant Editor, E. C. SANDERS 

Subscription Rates: United States and Mexico, $2.00 per year; Canada, $2.25 per year; Foreign, $2.50 per year; payable in 
advance. Remit by post-office or express money orders, bank checks or drafts, made payable to the General Electric Review, 
Schenectady, N. Y. 

Entered as second-class matter, March 26, 1912, at the post-office at Schenectady, N. Y., under the Act of March, 1879. 

VOL. XVIII- NO. 2 by r.eJir^E&iucLpany FEBRUARY, 1915 



Frontispiece 78 

Editorial : The Paths of Progress .... ... .79 

Developments in Electrical Apparatus During 1914 ... 80 

By John Liston 

The Absolute Zero 93 

By Dr. Saul Dushman 

The Towing Locomotives for the Panama Canal 1(11 

By C. W. Larson 

Electrophysics: Cathode Rays and their Properties 118 

By J. P. Minton 

The Selection of Railway Equipment 126 

By J. F. Layng 

A Short Method for Calculating the Starting Resistance for Shunt, Induction and Series 

Motors 131 

By B. W. Jones 

Application of the Coolidge Tube to Metallurgical Research 134 

By Dr. Wheeler P. Davey 

Effect of Altitude on the Spark-Over Voltages of Bushings, Leads and Insulators . . 137 

By F. W. Peek, Jr. 

The Lighting of Ships 143 

By L. C. Porter 

Practical Experience in the Operation of Electrical Machinery 146 

Current Transformer Failures; Heating and Sparking of Repulsion-Induction Motors; 
Excessive Pump Output. 

By E. C. Parham 

Notes on the Activities of the A. I. E. E 148 

From the Consulting Engineering Department of the General Electric Company . .152 




- E 


u i-4 



*" g. 

JS ^ 


j s 

it I 

c o (J U 

J I 

x u 

s » 

W ■= 

- £ c C 
u 5 o f 

a. c 

- V 

H S 


3 H 

c e 

< w 


-° 2 2. 
< x •* 


In our February issue we usually try to 
review the progress made during the year 
just past, so in this number we publish an 
article that outlines the progress made in the 
development of electrical apparatus during 
1914. Of necessity this review must be very 
incomplete as many developments are not 
learned till long after their inception and 
usually a considerable time elapses between 
the inception and application in actual 
practice. It will be noted that the progress 
cited is mostly in the nature of details of 
design and an increased capacity of ap- 

It would be a great mistake to surmise 
that this condition of affairs foretells any 
slowing down of progress in the electrical 
industry. Indeed, it rather lays emphasis 
on how far and how fast the art has advanced . 
We should note with interest and encourage- 
ment the almost daily invasion of electrical 
appliances to fields of work where formerly 
methods less up-to-date and less efficient were 
employed. The inherent characteristics of 
electrical apparatus and appliances seem 
bound to extend the use of electrical ma- 
chinery far beyond even its present enormous 
field, as we are verifying every day the fact 
that electricity furnishes the most flexible, 
reliable and efficient medium for trans- 
mitting energy from its source of origin to 
its many points of application. 

As we become more and more dependent 
on machinery for our economic progress, 
in just such a measure are we increasing our 
dependence for future developments on those 
who are perpetually increasing the efficiency 
of our electrical apparatus and rendering it 
more effective in its everyday applications. 
In reviewing progress we are apt to cite 
brilliant examples of discoveries, and to 
neglect giving due credit to those responsible 
for improvements in detail. In reality we, 
as a community, owe a tremendous debt to 
the "detail man" — the silent but perpetual 
worker, whose energy year in and year out is 
devoted to making improvements in details. 
These improvements in details constitute a 
host of inventions, many of them so small in 

themselves that they seldom merit the name, 
nevertheless we are becoming more and more 
dependent on them for our progress. 

The growth of the electrical generating unit 
to 35,000 kw. has only been made possible 
by an incessant study of, and improvement in, 
details; and the ever increasing potentials 
at which we can transmit our energy are due 
to the small rather than the large advances 
made in the art of design and construction. 

The advantage derived by the electrical 
industry as a whole and especially by the 
operating fraternity from this gradual advance, 
in distinction from a spasmodic development, 
have been great. To cite a specific example 
direct-current railway apparatus has de- 
veloped from the old standard of 600 volts 
through successive stages — 1200 volts, 1500 
volts, 2400 volts, up to the most recent 
3000-volt apparatus to be employed by the 
Chicago, Milwaukee & St. Paul Railway. 
As each step in advance was made new fields 
for electric traction were opened by the added 
economies to be secured — and old installa- 
tions in many cases adopted the higher 
potential apparatus as a means of effecting 
more economical operation on systems that 
had already been running for years. In no 
case was a wholesale discarding of electrical 
machinery, that still was capable of many 
years' good service, made necessary, which 
would have been the case had radically new 
development been substituted for a gradual 
improvement in details. 

So many of our modern developments are 
dependent upon the discovery and applica- 
tion of new materials, better suited to the 
severe conditions imposed by the constant 
demand for higher efficiency in weight, out- 
put, etc., than the materials formerly in use, 
that the research work done in this direction 
is constantly tying the industrial research 
laboratory closer and closer to the design 
office and the workshop. This phase of our 
industrial life has now reached a stage where 
the research laboratory must be looked upon 
as an indispensable factor in the modern 
manufacturing plant, if we are to keep abreast 
of the times and show a satisfactory rate of 
progress as each year passes. 




By John Liston 
Publication Bureau, General Electric Company 

It is often difficult clearly to comprehend the scope of the numerous minor changes effected in electrical 
manufacture during any given period, but a knowledge of the improvements thus made is essential in defining 
the yearly progress of the industry. In this article the author presents in a logical manner the improvements 
made in certain important classes of apparatus. — Editor. 

While some unique developments have 
characterized the progress made by the 
electrical industry during 1914, the general 
advance has consisted very largely of improve- 
ments in apparatus which had already 
attained relatively high efficiencies, both 
electrically and mechanically. 

Although man}' of the changes effected 
apparently concern only minor details of 
construction, their cumulative results show 
marked progress for the past year for the 
electrical art as a whole. Briefly stated, there 
has been achieved, refinement in design 
resulting in increased efficiencies for man)' 
classes of apparatus, economical concentra- 
tion of large energy values in single ma- 
chines, and a broadening of the field of appli- 
cation based on experiment and analysis of 
exhaustive operating data. 

In order adequately to represent the trend 
of design and manufacture, this review will 
refer briefly to certain specific cases which 
will serve to indicate the character and extent 

of recent improvements made in General 
Electric products; the data for the various 
sections being segregated under apparatus 

Steam Turbo-Generators 

Early in the year the first of the large 
horizontal Curtis steam turbine generator 
sets was placed in commercial service: It 
consists of a 20,000 kw. unit installed for the 
Commonwealth Edison Company of Chicago, 
and has already been in successful operation 
for almost a year. 

A still larger unit having an output of 
30,000 kw., 6600 volt, 25 cycles, at 1500 
r.p.m., operating normally under 185-pound 
steam pressure was placed in service by the 
New York Edison Company in November, 
1914, having been constructed and installed 
in less than a year ; a remarkably short period 
for a generating set of this capacity. The 
effective concentration of energy value 
achieved in the construction of this machine 

Fig. 1. 30,000-Kw. Steair. Turbo-Generator, New York Edison Co. 



is clearly indicated by the relatively small 
amount of space required for its installation; 
the overall dimensions being: Length, 57 
ft. 4 in.; width 19 ft. 8 in.; and height 14 
ft. 3 in. We hope to publish a detailed de- 
scriptive article covering this installation 
in an early issue of the General Electric 

A number of similar machines, ranging in 
capacity from 20,000 kw. to 35,000 kw., are 
on order, and several of these have been 
shipped, or are nearing completion in the 
Schenectady Works. It should be borne in 
mind that all of these large machines consist 
of a single generator direct connected to and 
mounted on the same bedplate with the 
turbine. They constitute the largest single 
generating units so far designed or con- 
structed by any manufacturer, and those 
already placed in service have without ex- 
ception established gratifying records in 
regard to reliability, steam economy and 
overall efficiency. 

The inherent simplicity and relatively 
compact arrangement of these large turbo- 
generators have made it possible to effect 
their installation in remarkably brief time 
when their great output is considered. As 
an example of this, a 12,500 kw. set was 
completely installed for the Toledo Railway 

Fig. 2. 100-Watt Steam Turbo-Generator for Steam 
Locomotive Lighting 

& Light Company, and placed in commercial 
service within fourteen days of its arrival at 

In striking contrast to the large machines 
referred to above is the diminutive turbo- 
generator developed during the year for sup- 
plying current for incandescent headlights 
and cab lights on steam locomotives. This 
set has a normal rating of 100 watts, 6 volts, 

at 3600 r.p.m., and a maximum continuous 
capacity of 140 watts. 

A steam pressure of about 90 pounds is 
maintained constantly by means of an auto- 
matic regulating inlet valve, and a safety 
pop valve is also provided. The turbine is 
a single-stage unit direct coupled to a direct- 
current compound wound generator, and by 
means of a differential brake magnet coil any 
fluctuations in the load are automatically 
compensated for so that constant voltage is 
maintained from no load to full load. 

This little self-regulating set has ample 
mechanical strength and has to date success- 
fully withstood severe practical service tests 
of more than six months duration, and it will 
undoubtedly have a wider field of application 
than that for which it was originally designed. 
Its overall dimensions are: Length, 23 J^ in.; 
width 15 in.; height, 14% in.; and its weight, 
130 pounds. 

Waterwheel Type Generators 

Conspicuous among the improvements for 
this class of apparatus is the suspension thrust 
bearing designed for vertical shaft type water- 
wheel generator sets, the bracket of which is 
rigidly supported by the generator stator, 
with the bearing carrying the entire weight 
imposed by rotor, waterwheel and water 
thrust. Among the larger machines for which 
these thrust bearings have been provided may 
be mentioned two 11,170-kv-a., 6600-volt, 
60-cycle sets, operating at 180 r.p.m.; three 
9000-kv-a., 12,000-volt, 40-cycle sets, operat- 
ing at 185 r.p.m.; four 10,000-kv-a. units, 
6600-volt, 60-cycle sets, operating at 200 
r.p.m. All of these generator sets are tested 
to withstand double normal speed, and the 
thrust bearings of the four groups referred to 
sustain respectively aggregate weights per 
unit of 75, 77, 100 and 175 tons. 

An indication of a recent tendency in hydro- 
electric development, brought about pri- 
marily through improved efficiency in water- 
wheels, is the use of generators of relatively 
small capacity and low speed which have 
rendered it possible effectively to utilise 
numerous low head water powers which here- 
tofore could not be economically developed. 
Among the machines which have been con- 
structed to meet these conditions during the 
past year, with a rating lower than 1000 
kv-a., are generators having rated capacities 
of 600 kv-a., at 48.5 r.p.m., down to 150 
kv-a. at 180 r.p.m. Slightly larger units 
have been utilized at relatively low speeds 
and these may be typified by reference to 



six 2000-kv-a., 6600-volt, 25-cycle sets, which 
are designed for operation at 68.5 r.p.m. 

Among the larger sets mav be mentioned 
the two 12,000-kv-a., 6600-volt, 60-cycle 

Fig. 3. 12,000-Kv-a. Waterwheel-Driven Generators, 
Light & Power Co.. Grace, Utah 


vertical shaft waterwheel-driven generators 
placed in operation by the Utah Power & 
Light Company at its main generating station 
at Grace, Utah. While these machines have 

been exceeded in capacity by generators of a 
similar type previously installed, they have 
been designed and constructed for operation 
at the highest speed at which sets of this 
capacity have as yet been called on to operate ; 
i.e., 514 r.p.m. 

In order to obviate the destructive effects 
frequently produced by corona in high 
potential generators and synchronous motors, 
there has been devised a corona shield which 
has proven thoroughly practical in operation 
and has been used to a constantly increasing 
extent during the past year. It consists of a 
layer of tinfoil placed over the ordinary 
insulation and covers that part of the coil 
which is enclosed by the slot, extending far 
enough beyond the slot to give ample room 
for protecting, with tape and varnished cam- 
bric, the projecting ends of the tinfoil covering. 
The corona shields are finally connected. 
by thin copper strips, with the stator lamina- 
tions, through which they are effectually 

A number of direct-current generators of 
exceptional capacity, designed for water- 
wheel drive, have also been constructed 
during the year, and at present work is near- 
ing completion on a lot of 1 1 horizontal shaft 
direct-current machines of this type, each 
having a rated output of 5200 kw., 520 volts, 
at 170 r.p.m. These exceed in size any gen- 
erators of their type previously built. 

Gas-Engine Driven Generators 

Due to improvements in design and regu- 
lation, the past year has witnessed consider- 

Fig. 4. 6250-Kv-a. Waterwheel-Driven Generator Showing Construction of 
Suspension Thrust Bearing Bracket 



able advance in the use of 
60-cycle gas-engine driven 
generators, and at the pres- 
ent time there are nearing 
completion three units of 
1390 kv-a. capacity, 2300 
volts, 60 cycles, arranged 
for operation at 116 r.p.m. 
They will be utilized by 
the Monongahela Traction 
Company of Fairmont, West 
Virginia, and are the largest 
60-cycle generators designed 
for gas-engine drive. Other 
and larger units had, how- 
ever, been constructed prior 
to 1914 for 25-cycle opera- 
tion; the rating for these 
machines, which were built 
for the Bethlehem Steel 
Company, being 3125 kv-a. 
An equitable basis of 
guarantee for the parallel 
operation of both gas-engine 
and steam-engine driven 
generators has been devel- 

Fig. 6. 

5. 5200-Kw., 250- Volt Direct Current Generator 
for Waterwheel Drive 

the General Electric Company 
generally accepted, should prove 

oped by 

which, if c 

of considerable value to the builders of engines 

and generators, and to the operator. 

2250-Kw. Synchronous Converter with Synchronous 
Booster, Boston Edison Co. 

This guarantee is based largely on the 
determination of the natural period of the 
generator in relation to the various character- 
istics of the complete unit, including flywheel 
and the operating features of engines and 
governors, and presents in a logical manner 
the data necessary for an accurate pre- 
determination of results. 

Synchronous Converters 

The advance in this class of apparatus has 
been marked chiefly by the increased unit 
capacity of 60-cycle machines produced; a 
representative installation consisting of two 
2250-kw., 225/275-volt converters equipped 
with synchronous boosters has been con- 
structed for the Boston Edison Company. 
These machines have approximately 50 per 
cent greater capacity than any 60-cycle 
machines of this type heretofore developed, 
and their successful operation has resulted in 
the adoption of still larger units. 

As evidence of this a single order was 
received for eighteen 500-volt, 60-cycle syn- 
chronous converters with commutating poles, 
having an output of 2500 kw. each ; an aggre- 
gate rating of 45,000 kw. They are to date 
the largest 60-cycle synchronous converters. 
This record order has already been partially 
filled, and the machines will be installed at 
Messine, N. Y., for the Aluminum Company 
of Americp.. 



The equipment of synchronous converters 
with synchronous boosters which are integral 
parts of the complete machine, together with 
an arrangement for automatic control, con- 
stitutes a most important improvement in 
synchronous converter operation and insures 
a positive and automatic adjustment of the 
direct-current voltage. This is accomplished 
through control of the field excitation of 
synchronous converters provided with com- 
mutating poles, and insures correct excitation 
at all loads and voltages. 

Synchronous Condensers 

A horizontal shaft 6000-kv-a., 50-cycle, 
500-r.p.m. synchronous condenser constructed 
for the Southern California Edison Company 
of Los Angeles is of unusual interest due to 
the fact that it was designed for operation 
on a 16,500-volt line; no machines of this 
type having previously been built for poten- 
tials exceeding 6600 volts. 

In addition to the precautions necessary for 
insulation against this unusual potential, 
special efforts were made in designing the 
machine to minimize the losses, and as a 
result there was produced a synchronous 
condenser of vers- high efficiency which is 
utilized for power-factor correction. It is 
self-starting by means of a compensator and 
requires less than half its rated kilovolt- 
amperes for starting. 

Phase Advancers 

While synchronous condensers are ordi- 
narily applied for improving the power-factor 
of a system, the phase advancer is designed 
primarily for improving the power-factor of 
individual induction motors, although in 
special cases it is capable of wider application. 
This machine, which has been made com- 
mercially practicable within the past year, 
is described in the June, 1914, General 
Electric Review, but its characteristics 
may be briefly outlined as follows: 

The phase advancer stands in the same 
relation to an induction motor as an exciter 
does to a synchronous motor. However, 
for the induction motor, continuous current 
can not be used for the magnetizing current 
in the secondary because the motor slips 
under load. The magnetizing current must 
be a polyphase current of low frequency 
which corresponds in each instance to the 
slip of the induction motor. 

The phase advancer consists of a con- 
tinuous-current drum armature with a com- 
mutator having three brush studs per pair 

of poles displaced relatively to one another 
by 120 electrical degrees. The stator merely 
consists of a frame with the laminations 
assembled but having no slots or windings. 

The phase advancer is direct connected to a 
small squirrel cage constant speed induction 
motor. The power necessary to drive the 
phase advancer is only that required to supply 
the friction windage and hysteresis losses and 
is therefore comparatively small, i.e., about 
one h.p. for a 600-h.p. 2200-volt induction 
motor. The copper losses are provided by 
the main induction motor rotor. 

Frequency Changers 

A notable frequency changer set was 
recently installed to interconnect the Boston 
Edison and Boston Elevated systems. This 
is a horizontal shaft set, the 60-cycle unit 
being rated at 9500 kv-a., 13,800 volts, and 
the 25-cycle unit rated at 9000 kv-a., 13,200 
volts; it operates at 300 r.p.m. and is rever- 
sible. It is totally enclosed and provided 
with inlet for external air supply which dis- 
charges into the station. One frame is ad- 
justable so that if the equipment is duplicated 

Fig. 7. 7500-Kv-a., 12.000 24,000 Y, Three-Phase 
Water-Cooled Core Type Transformer 

both sets may be arranged equally to share 
the load, and in order to facilitate inspection 
or to make repairs each frame is arranged 
to move on steel rollers parallel to the shaft. 
This is the largest frequency changer set 



produced by the General Electric Company, 
and is probably the largest machine of this 
type in service today. 


Prior to 1914 the largest core type trans- 
formers produced by the General Electric 
Company did not exceed 2000 kv-a. in rated 
capacity, but during the year the maximum 
rating was carried up to 7500 kv-a. 

The maximum rating of single-phase water- 
cooled shell type transformers has also been 
increased by the construction of four units of 
8333-kv-a. capacity. 

There has been a marked reduction in 
interruption to service in transformers of 
recent design as they are now capable of 
withstanding momentary short circuits under 
sustained primary voltage without injury to 
the coils. This has been accomplished 
largely through changes in the grouping of the 
coils and working to higher inherent re- 

A feature of unusual interest for the year 
is the development of a combination trans- 

of normal load without the circulation of 
water and without exceeding its specified tem- 
perature rise. On the other hand, this trans- 
former may be designed for normal operation 
as a self-cooled unit, and be provided with 

Fig. 8. 8333-Kv-a. Single-Phase Water-Cooled Shell 
Type Transformer 

former which may be operated either self- 
cooled or water-cooled. It may be designed 
for normal operation with water circulated 
through the cooling coils, in which case it 
may also be safely operated at 50 per cent 

Fig. 9. 1250-Kv-a. Combination Self-Cooled-Water- 
Cooled Outdoor Transformer 

the necessary cooling coils which, when 
utilized, permit operation efficiently at 50 
per cent above the normal capacity. The 
economical advantages of such transformers 
are obvious, expecially for localities where 
the purchasing rate of water is high and the 
transformer is fully loaded only part of the 
time. This design also provides a factor of 
safety in case of interruption in the water 
supply, in which event the apparatus may 
still be operated at partial load instead of 
being shut down. 

Feeder Regulators 

Early in 1914 a single order for 100 feeder 
regulators was received from the Common- 
wealth Edison Company, Chicago, 111. These 
regulators are rated at 36 kv-a., 60 cycle, 150 
amp., and are of the automatically operated 
induction type, designed for 2400-volt pri- 
mary and 240-volt secondary. When placed 
in service they are utilized for maintaining 
constant voltage on alternating-current 
feeders having an aggregate capacity of 
36,000 kv-a. Prior to the placing of this, 
the largest single order for this type of ap- 



paratus, the Commonwealth Edison Company 
had already installed approximately 300 
General Electric regulators for similar ser- 
vice, making a notable aggregate equipment 
of 400 units of this type. 


With the steady growth in the size of power 
plants and distribution systems, there has 
arisen among operators a fuller conception 
of the practical value of providing ample 
protective equipment, such as voltage reduc- 
ing devices for cutting down the station volt- 
age under short circuiting conditions, arcing 
ground suppressors for short circuiting arcs 
caused by grounded phases on delta connected 
transmission systems, and the recently de- 
veloped and improved current limiting react- 
ances. A fuller appreciation of such devices 
is amply attested by a continual increase in 
the number of propositions for power station 
equipment which include this class of ap- 

The reactance developed by the General 
Electric Company has been improved to a 
considerable extent during the past year and 
the changes made have been based largely 
on the experience gained in numerous applica- 
tions of the earlier types utilized for the pro- 
tection of feeder lines. In order to facilitate 
calculations on the equipment necessary to 
meet the requirements of power systems hav- 
ing widely varying operating 
conditions and capacities, these 
reactances are now made in 
three distinct forms. 

Electric Railways 

The decision of the Chicago, 
.Milwaukee & St. Paul Rail- 
way Company to electrify its 
mountain grade divisions in 
Montana marked one of the 
most important steps ever 
taken in steam road electri- 
fication. The order for high- 
voltage direct-current electrical 
equipment placed with the General Electric 
Company includes nine freight and three 
passenger locomotives, weighing approxi- 
mately 260 tons each, all equipped for regen- 
erative braking; 10 three-unit synchronous 
motor-generator sets with transformers; and 
switching apparatus for the equipment of 
four substations totaling 17,000 kw. in ca- 
pacity. Overhead line material is also in- 
cluded for the initial electrification of 113 
miles, or 168 miles on a single track basis. 

The railroad company has plans under way 
for the electrical operation of the entire 440 
miles of main line transcontinental road 
between Avery, Idaho and Harlowton, Mon- 

In the selection of 3000 volts direct current 
as the operating voltage, this road was 
doubtless influenced to a large extent by the 
attractive performance of the 2400-volt 
direct-current equipment of the Butte, Ana- 
conda & Pacific Railway, t 

An interesting railway is being constructed 
by the Bethlehem-Chile Mines Company in 
Tofo, Chile. This road will be used for con- 
veying iron ore from the mines about 2000 
feet above the sea level, a distance of about, 
15 miles to the Port of Cruz Grande on the 
coast. The equipment of this road will 
include three 110-ton, 2400-volt direct-cur- 
rent electric locomotives which will be sup- 
plied by two three-unit 1000-kw., 2400-volt 
synchronous motor-generator sets. This sub- 
station will be fed over a 22,000-volt trans- 
mission line from a main power house which 
will contain three 3500-kv-a. and one 300- 
kv-a. Curtis steam turbines. 

The average grade on this road is about 
three per cent and the locomotives are to be 
equipped for regenerative control feeding 
power back into the system on the down 

Work is proceeding rapidly on two other 

Fig. 10. Motor-Generator Set for Canadian Northern Railway. 2100-H.P., 
11,000-Volt MotorlDriving Two 750-Kw., 1200-Volt DC. Gener- 
ators Connectedlin Series~for~2400-Volt Railway Service 

2400-volt railway electrifications, the Michi- 
gan Railway Company and the Canadian 
Northern Railway, which are expected to be 
in commercial operation early in 1915. 

Another important endorsement of the 
high voltage direct-current system is the 
decision of the Ontario Hydro-Electric Power 
Company to employ 1500 volts direct current 
for the electrification of the London & Port 

* See General Electric Review. Nov.. 1914. 
t See General Electric Review, Jan.. 1915. 


Stanley Railway.* Orders have been placed 
for the initial rolling stock, including three 
60-ton electric locomotives, five four-motor 
multiple unit cars, and four trail cars. This 
road is about 24 miles long and connects 
Port Stanley on Lake Erie with London, 
Ontario. The electrification of this 
steam road division is the beginning 
of an extensive system owned and 
operated by the municipalities in 
this section. 

High-voltage, direct-current 
equipment has also been ordered 
during the year for a number of other 
interurban railways, including the 
following: Chicago, Milwaukee & 
St. Paul, (Great Falls Electrifica- 
tion), 1500 volts; Imperial Railways 
of Japan, 1200 volts; Toronto Sub- 
urban Railway, Canada, 1500 volts; 
Willamette Vallev Southern Rail- 
way, 1200 volts. 

The Pacific Electric Railway has 
ordered 96 ventilated motors for 
new cars on the Los Angeles-San 
Bernardino-Riverside division. 

The principle of ventilation as employed 
on all modern General Electric motors "has 
been adopted by many railways in all parts 
of the world. Motors of this type have been 
selected by the New York Municipal Railway 
Company for the operation of 200 new cars 
in the new Brooklyn Subway, and 334 motors 
of a similar type have also been ordered by the 
Northwestern Elevated Railway of Chicago. 

cars. Four-hundred of these motors are being 
placed in service on the Pittsburgh Railways. 

Mine Locomotives 

All of the mine locomotives manufactured 
by the General Electric Company in 1914 were 

Fig. 11. GE-248-A Railway Motor Adopted for New 
York Municipal Railways 

The Chicago Surface lines are using 200 
GE-242 ventilated motors which were ordered 
during the year, and delivery is being made 
on an additional order for 456 of these motors. 
The GE-247 is a new ventilated type railway 
motor designed for 24-inch wheel low floor city 

* See General Electric Review Jan., 1915. 

Fig. 12. 50-Ton 1200-Volt Locomotive for Willamette Valley Southern Railway 

provided with commutating pole motors and 
ball bearings as standard equipment, and 
the operating records of those placed in service 
during the year show that these improve- 
ments have reduced the number of interrup- 
tions to service and have resulted in de- 
creased maintenance costs. 

The increasing output of many mines has 
rendered it necessary to equip them with 
locomotives of relatively large capacity, 
capable of handling heavy trips over steep 
grades and for long hauls. For this class of 
service there have been built a number of 
three-motor, 15- and 20-ton locomotives. 
The 20-ton unit combines some unusual 
features in design and construction: The 
body is made of rolled steel, each side frame 
being cut from a solid rolled steel slab, while 
steel slabs in conjunction with steel channels 
are used for the end frames. The three 
driving motors are each rated at 85 h.p. and 
are of the split frame type. These particular 
locomotives were built for 42-inch gauge, but 
the same construction and capacity can be 
utilized for a minimum of 36-inch gauge. 

Up-to-date practice in haulage locomotives 
may be represented by reference to the con- 
structive features of a typical 16-ton single- 
truck three-motor unit. In this, the latest 
type of industrial locomotive, the truck frame 
is built of steel throughout, both the sides and 
ends being cut from single pieces of solid 
slab. The platform is built of steel channels 



and plates, and the cab of steel sheets. It is 
a standard gauge machine and, in so far as 
possible, all details have been developed 
along the lines of standard railway practice, 
the wheels, axles, journal boxes, brake beams, 
brake shoes and couplings being all in accord- 

Fig. 13. 20-Ton, Three-Motor Mine Locomotive, 42-In. Gauge 

ance with MCB requirements. It is driven 
by two 60-h.p., 500-volt motors and equipped 
with straight air brakes. 

An interesting type of locomotive has also 
been constructed for service at the mines of 
the Braden Copper Company in Chile, S. A. 
It is a 25-ton double-truck machine for 30- 
inch gauge, and has an overall height of only 
seven and one-half feet. The four driving 
motors are each rated at 45 h.p., 250 volts, 
multiple unit control and automatic air 
brakes being also included in the equipment. 
It is probable that this locomotive is the 
heaviest and narrowest gauge, and has the 
lowest overall height of any machine of this 
type ever built. 

There has been a definite increased demand 
for the storage battery type of locomotive 
for gathering work, as it has been demon- 
strated that in this service each locomotive 
will effectively displace at least two or three 
mules. Heavy units are not as a rule re- 
quired and the locomotives of this class so 
far provided have been rated at from three to 
seven tons. Most of these are of the straight 
storage battery type, but a limited number 
have, in addition, been equipped so that they 
can operate from a trolley wire when in the 
main headings of a mine. The advantages 
of this arrangement are obvious in that bv 
means of a small self-contained motor-gen- 
erator set, the battery may be automaticallv 
charged while the locomotive is running on 
the trolley. When the locomotive is working 
in the rooms, gathering the cars, a varying 
percentage of the battery charge will be con- 
sumed, but as soon as the locomotive is again 
operated on the trolley, these losses are auto- 

matically compensated for and with this dual 
system of operation the battery need never 
be entirely discharged and if space limitations 
are severe it permits the use of a smaller 
battery than would otherwise be necessary. 
A representative machine of this type has 
been in operation in a West Virginia mine for 
a period of about four months. It runs on a 
42-inch gauge track and its overall height 
does not exceed 30 inches. 

Mine Hoists 

The largest induction motor shaft hoist 
equipment in America was placed in operation 
in November, 1914, at Lansford, Pa., for the 
Lehigh Coal & Navigation Company. The 
driving motor is rated at 750 h.p., 300 r.p.m., 
three-phase, 25 cycle, and drives through a 
single reduction gear. 

Positive control of the hoisting speed is 
secured by means of an improved type of 
liquid rheostat and high tension air break 
contactors; the motor circuit being 2300 
volts. This hoist serves a 600-ft. vertical 
shaft hoisting 11,500 pounds per trip at the 
rate of 90 trips per hour, with a maximum 
rope speed of approximately 1600 feet per 

The liquid rheostat referred to above was 
developed primarily for mine hoist service 
and insures safe operation at quick reversal. 
It employs two sets of fixed electrodes at 
different elevations; one set being widely 
spaced, while the other set has large electrode 
areas and has small spacing in order to obtain 
a very low final slip. The two sets of elec- 
trodes are connected in parallel after the 
electrolyte has reached a certain level corre- 
sponding to a predetermined decrease in rotor 

Fig. 14. Combination Storage Battery and Trolley Type Mine 

Locomotive with Platform Removed to Show 

Internal Arrangement 

voltage. All parts of the rheostat itself are 
stationary, thus insuring absolute reliability; 
the electrolyte level being varied through the 
operation of a movable weir and a small 
motor-driven pump. 



Steel Mills 

The tendency toward the exclusive use of 
electricity for all power application in modern 
rolling mills is indicated by the equipment 
selected by the Bethlehem Steel Company for 
its new plant at South Bethlehem. In equip- 
ping the new buildings no steam drive or 
steam auxiliaries have been provided. The 
electrical energy is generated with gas engine 
drive and for power application three-phase, 
25-cycle, 6600-volt induction motors have 
been used throughout. In this new plant 

Electric Furnaces 

The fact that the electric furnace offers a 
compact, reliable and economical method of 
manufacturing crucible quality steel has 
now become more generally recognized among 
iron and steel founders and, in consequence, 
there has been an appreciable increase in the 
number and size of the equipments recently 
installed or in process of construction, and a 
concomitant improvement in details tending 
toward improved efficiency. 

Perhaps the most striking advance has been 

Fig. 15. 750-H.P. Induction Motor Driving Mine Hoist 

there are General Electric motors ranging 
from 350 h.p. to 3000 h.p. both of the single- 
speed and two-speed pole changing type, the 
aggregate rating being approximately 12,000 

In order to provide speed control for in- 
duction motors which will meet the variable 
load requirements of rolling mills a speed 
regulating set applicable for this class of 
work has been developed, and during the 
year has been practically applied by the 
American Iron & Steel Company, Penn- 
sylvania Steel Company, Forged Steel Wheel 
Company and Union Rolling Mills. These 
speed regulating sets enable the induction 
motor to carry varying loads at constant 
speed, giving it for all practical purposes the 
speed characteristics of the shunt wound 
direct-current motor, while at the same time 
retaining the simple and strong mechanical 
features of the induction motor. 

in the induction type of furnace as, prior to 
1914, the largest unit of this type in the 
United States had a capacity of only two 
tons, whereas during the year this was carried 
to 20 tons, two units having been completed. 
This 20-ton furnace is of the two-ring type 
and in operation utilizes single-phase current 
of five-cycle frequency at 5000 volts, and it 
has been necessary to supply a special motor- 
generator set in connection with it, consisting 
of a two-pole single-phase generator having 
an output of 4000 kv-a. at 5000 volts, five 
cycles, which is direct driven by a three-phase, 
25-cycle, 2300-volt synchronous motor. 

The exceptional size of the furnace, which 
is the largest of any type in the United States, 
used for refining steel, is best illustrated by 
reference to the core and coils, which ele- 
ments for each furnace have a weight of 
approximately 60 tons. In operation the 
furnace rings are charged with molten metal, 



every part of which is thereby subjected to 
intense, uniform and positively controlled 
heat, and is then poured off after a treatment 
lasting from 60 to 90 minutes. 

For the arc type of electric furnace special 
forms of transformers and auxiliary equip- 
ment have been designed, together with a 
reliable system of automatic control which is 
particularly interesting to the practical oper- 
ator, in that, except for a short period after 
starting the furnace, a constant power input 
is maintained at such a value as may be pre- 
determined by the operator. 

The resistance type of furnace, which 
utilizes heat generated by passing the electric 
current through a resistor composed of 

Fig. 16. 

Arrangement of Core and Coils for 20-Ton, 
Two-Ring Induction Furnace 

foundry coke, with auxiliary heating from a 
carborundum arch which radiates heat down- 
ward on the charge, has also been provided 
with a simple current relay control which 
insures the maintenance of a constant tem- 
perature over a range of approximately 600 
to 1300 deg. C. 

Switching Apparatus 

There are numerous localities remote from 
low-voltage distribution, but accessible to 
high-voltage transmission lines, where the 
small rural substation can be economically 
utilized for the distribution of electrical 
energy in capacities as low as 3 kv-a. Under 
proper conditions a market of this kind offers 
the operating company a sound financial 
basis for sendee, providing that the equip- 
ment can be installed at a reasonable price, 
and can be depended upon to operate with 
low maintenance expense. The great extent 
of the field for this class of electrical equip- 

ment is indicated by numerous successful 
substations in small towns, farms, mines and 
quarries, pumping outfits, small isolated 
manufacturing plants and various contracting 
and construction jobs. 

Experience has shown that to avoid inter- 
ruptions to service the equipment supplied 
to meet the operating conditions indicated 
above must be proof against damage from the 
weather and, in order to provide this protec- 
tion, all operating parts of switching appara- 
tus supplied by the General Electric Com- 
pany, which are liable to rust, are given a 
very effective protective treatment. 

While considerable work has been ac- 
complished in the effective equipment of 

Fig. 17. Resistance Type Annealing Furnace with Control 

Panel — Later Forms of this Furnace are Sheathed 

with Cast Iron Plates 

rural substations prior to 1914, the activities 
of the past year have been along the line of 
standardizing apparatus suitable for use in 
connection with various transmission volt- 
ages, and in improving or redesigning standard 
apparatus previously utilized. The complete 
standardized line now available has been 
proven reliable in service and is designed to 
minimize danger to the equipment in case of 
disturbance on the main line, and while it 
is proof against injury to itself or other 
apparatus, it does not require skilled attend- 
ance. Furthermore, it is very largely semi- 
portable, so that it can be installed or re- 
moved promptly and economically, and the 
fact that the entire line .has been standard- 
ized permits an accurate predetermination 
of the cost of a rural substation when the 
operating conditions and service required 
are known. It also makes it possible to secure 
for the small community many of the econo- 
mies inherent in high tension transmission. 



Fig. 18. 

2500-Volt, Three-Phase Switch House 
for Outdoor Installation 

A new type of solenoid-operated manhole 
oil switch has been developed which is entirely 
self-contained, and gives the advantage of 
remote control in that the switch may be 
operated from a distance with absolute 
reliability, even if the switch compartment is 
completely flooded. The compact arrange- 
ment and water-tight construction of this 
oil switch render it specially valuable for 
manhole service, or for use in other locations 
liable to flood. 

A number of notably large circuit breakers 
have recentlv been constructed, the maximum 

Fig. 19. 2500-Volt, Three-Phase Switch House Showing 
Inside View with Meter Panel Removed 

capacity provided for being 20,000 amperes. 
This type of breaker is operated by a single- 
coil solenoid with the usual automatic over- 
load trip, and has, in addition, a shunt trip 
coil plunger which acts directly on the cir- 
cuit breaker locking latch, instead of on a 
trip on the solenoid. 

A new and ingenious electrostatic syn- 
chronism indicator has been developed. The 
instrument case resembles an ordinary round 
pattern switchboard instrument, and inside 
of this are receptacles for holding three 
special glowers which project through holes 
in the case cover. All connections from the 
line to the device are made through con- 
densers which consist of suspension insulators 

, ™_ ... 


Fig. 20. Arrangement of Switching Apparatus and Three Single - 

Phase Transformers for Supplying 150 kv-a. at 2300 Volts 

from a Three-Phase 35,000- Volt Transmission Line 


Fig. 21. Electrostatic Synchronism Indicator 



having an insulation equal to that used on the 

Normally, the glowers have the appearance 
of ordinary spherical frosted incandescent 
lamp bulbs, but when there is a proper 
difference of potential across the terminals, 
they glow with a reddish hue, due to the use 
of a special gas. 

crease in size and modifications in the shape 
of the lamp which were found advisable due 
to the concentration in a single incandescent 
unit of the large candle-powers which the 
high efficiency type of lamp permitted, but 
largely because of the increased temperatures 
experienced in their operation. 

During the past year these lamps have been 
successfully adopted for street lighting in a 
number of cities, and for series operation 
they have been provided with a new type of 
compensator having efficiencies of from 93 
to 95 per cent with power-factors of from 97 
to 98 per cent. 

There has also been developed by the 
General Electric Company a prismatic re- 
fractor which, in combination with suitable 

Fig. 22. Compensator Type Incandescent Lighting Unit 

Equipped with Concentric Reflector and 

Prismatic Refractor 


— s— ^_ 



>-£*"* Dlt ' TVk *^^\^\^^)( 


v\ 7-i 



Fig. 23. Characteristic Distribution Curve of the Above 
Unit with 600 c-p. Mazda Series Lamp 

The instrument can be operated on a line 
having a pressure as low as 13,200 volts, and 
can be made suitable for practically any 
voltage above this by simply cutting in the 
proper number of insulators. 


The commercial application of the high 
efficiency mazda lamp involved the design 
of a complete new line of fixtures. These 
were rendered necessary, partlv bv the in- 

Fig. 24. Luminous Arc Lamp with Prismatic Refractor 

reflectors, insures a more effective control of 
the light distribution than any type of globe 
yet developed. 

For street car headlights a new line of con- 
centrated filament incandescent lamp has 
been provided, and a large percentage of 
those now in service are equipped with a new 
and highly efficient form of glass parabolic 

The improved efficiencies which have been 
obtained during the year in luminous arc 
lamps have been due very largely to the pro- 
duction of an electrode which, for a given 
current consumption, produces from 30 to 
50 per cent more light than any electrode 
previously utilized. 

This has been accomplished through ex- 
haustive research work resulting in new 
electrode compositions, containing a larger 
proportion of titanium than older electrodes. 
The use of this element in suitable combina- 
tion enables this type of arc lamp to give 



the highest illumination efficiencies yet ob- 
tained by any commercial lamp. In addition 
to this the arc lamp mechanism has been 
simplified and the light distribution improved 
by the adoption of a prismatic refractor 
similar in principle to that designed for the 
high efficiency mazda lamp, but differing 
from it in form. 

These cumulative improvements have re- 
sulted in the production of a new line of 5- 
ampere luminous arc lamps which give 
practically the same illuminating values 
formerly obtained with 6.6-ampere lamps. A 
5-ampere series rectifier is also available for 
operation in connection with the new lamps. 

The vast number and varied character of 
the developments in the modern electrical 
industries renders it exceedingly difficult 
to give in a necessarily limited article, a truly 
comprehensive description of the progress 
made in any year, but in the foregoing the 
writer has endeavored to show the general 
trend of the changes inaugurated in the manu- 
facture of electrical apparatus, by reference 
to a limited number of specific examples and, 
in conclusion, it may be reiterated that most 
of the equipments cited have already been 
subjected to the stresses of commercial ser- 
vice and have successfully withstood the 
pragmatic test. 


Part I. 

By Dr. Saul Dushman 

Research Laboratory, General Electric Company 

During the last three or four years a large number of important investigations have been carried out on the 
properties of substances at extremely low temperatures. The results obtained have been intensely interesting, 
both from a practical and theoretical point of view. In the first part of the paper the author discusses the 
logical foundations of our present temperature scale and the various methods that have been used to attain 
extremely low temperatures; while in the next issue he will deal with the behavior of different substances at 
low temperatures and point out the important bearing of these investigations. The original conception of tem- 
perature was simply that of denoting the state of heat or cold of a body. Subsequently, the necessity arose for 
quantitatively comparing different states of heat or cold. — Editor. 

Conception of Temperature 

Early conceptions of temperature, heat, 
cold, and quantity of heat were very confused. 
There was a great deal of groping in the dark 
before the idea of measuring heat quantita- 
tively was arrived at and the difference between 
quantity of heat and temperature was under- 
stood. It was generally known that the 
volume of a body altered with its state of 
heat or cold and so there followed the idea 
of using a volume of a mercury or a gas 
column, placed in contact with the body, as 
a mode of determining its temperature. 

Having noticed that there is no change 
in volume at the melting point of ice and 
the boiling point of water (at constant 
atmospheric pressure) it was also agreed to 
use these two fixed points for the graduation 
of thermometers. 

There arose, however, the necessity of 
indicating temperatures above and below 
these two fixed points, and the question also 
arose as to the manner in which the scale 
between the two fixed points should be 

On comparing the expansion of different 
substances it was observed that the expansion 
in most cases is approximately linear. Thus 
when we use as thermometric substance a 
mercury column and divide the volume 
between the two fixed points into one hundred 
equal divisions, we find that whether we take 
air, alcohol, or glass, each of these substances 
has an approximately constant coefficient 
of expansion throughout the same range of 
temperatures. It thus became possible to 
extend the scale of temperatures below the 
freezing point of water and above its boiling 

A further generalization was observed. 

All gases expand about — — of their volume 

at the melting point of ice when the tem- 
perature is raised to that of boiling water. 
Here then we have a property which is 
independent of the nature of the substance 
used. What could be more natural than the 
decision to use a gas as standard thermo- 
metric substance? 



Gas Thermometry 

It is necessary to distinguish in this con- 
nection between the gas thermometer at 
constant pressure and that at constant 
volume. Denoting the temperature on the 
Centigrade scale by t, and the corresponding 
pressure and volume by P t and V respec- 
tively, the temperature is denned as follows: 

t = 273 

at constant volume 

at constant pressure 

where P and V' denote the pressure and 

volume respectively at the freezing point of 



Now if we plot / as abscissa and ■=■ as 

l o 

ordinate, we find that at t = 273, PtjP be- 

V t 

comes equal to zero; similarly j? becomes 

equal to zero at t=273. In other words, on 
the constant volume thermometer, the pres- 
sure vanishes at a temperature of — 273 deg. 
C, while on the constant pressure ther- 
mometer, the volume vanishes at this tem- 
perature. Here then we apparently reach a 
non ultra plus in the region of low tem- 
peratures. We might, therefore, be justified 
in designating this lowest temperature as an 
absolute zero. 

At first glance, the conclusion based on 
the above considerations that there must 
exist a lower limit of temperatures might 
seem rather arbitrary. Why choose -273 
deg. C. any more than -5500 deg. C, which 
corresponds to the temperature at which a 
volume of mercury would vanish? (The 

coefficient of expansion of mercury is --,.,. 

per degree Centigrade.) When it is, however, 
considered that all substances in the gaseous 
state exhibit practically the same coefficient 
of expansion, and furthermore that gases 
probably represent the simplest state in 
which matter can be obtained — when these 
facts are duly considered, it is seen that the 
choice of -273 deg. C. as the absolute zero 
is not so arbitrary. 

There is, however, a much more cogent 
reason for concluding that an absolute zero 
actually exists and that it coincides with the 
absolute zero as defined on the ideal gas 

Absolute Scale of Temperature 

It was first pointed out by Lord Kelvin 

that the scale of the ideal gas thermometer 

coincides with another scale of temperatures 
which can be based upon the second law of 
thermodynamics and is therefore independent 
of the properties of any particular substance. 

In its most general form this law states 
that for the conversion of energy in the form 
of heat into mechanical energy there is 
required a difference in temperature; and the 
maximum fraction of the total heat energy 
at any given temperature that is convertible 
into work depends upon the available tem- 
perature drop only and not upon the nature 
of the engine used for the operation. It is 
for this reason that we cannot withdraw heat 
from the ocean and convert it into mechanical 
work. The steam engine, as well known, is 
a direct application of the above principle. 
The greater the difference in temperatures of 
boiler and condenser, the greater the effi- 

Let us denote by Q\ the amount of heat 
absorbed at the higher temperature, Q u by 
any sort of reversible engine that is capable 
of converting heat into work. For our 
present purpose we do not need to worry 
about the exact thermometric scale which 
we adopt to measure 0. Let 2 denote the 
temperature of condenser and Q* the heat 
given out by the engine at this temperature. 
The difference 0\ — Qi corresponds to the 
amount' of heat converted into work, and the 
fraction (Qi — Qi)!Qi measures, therefore, the 
efficiency of the process. 

According to the second law of ther- 
modynamics, this efficiency depends upon the 
temperatures 0i and 2 , that is, upon the tem- 
peratures of the hot and cold reservoirs. We 
can assign to B\ and 02 such values that they 
represent the ratio of the quantities of heat 
Q\ and Q 2 ; that is, we write 

Q* = 02 
01 Ox 

This manner of reckoning temperatures 
immediately leads us to the notion of an 
absolute zero of temperature, for the equation 
can be written in the form 

At 02 = 0, all the heat Qi taken in from the 
hot reservoir will be converted into work, 
and since we cannot imagine any better 
efficiency than this, it- is impossible for 0? 
to be negative. "Thus 2 = O is the lowest 
temperature conceivable. The zero on this 
scale is consequently an absolute zero of 
temperature independent of the properties 
of any particular substance, for when the 



efficiency of one reversible engine is unity, 
the efficiency of every other reversible engine 
working between the same source and con- 
denser will also be unity, and hence if is 
zero for one substance, it will also be zero 
for every other. This zero is therefore ab- 

We are still at liberty to choose the size 
of a degree on this scale. If we choose the 
new scale so that there may be 100 degrees 
between the freezing and boiling points of 
water, we find that the absolute zero is 273 
degrees below the freezing point of water. 
Definition of Ideal Gas 

The scale of an ideal gas thermometer is 
therefore identical with the absolute scale 
defined in the above manner. In this con- 
nection we may define an ideal gas as "one 
which follows Boyle's law and in which a 
free expansion, with no external work, would 
cause no change in temperature * * * * No 
real gas satisfies these conditions exactly, 
but all the common thermometric gases, as 
they are used in gas thermometers, do satisfy 
them approximately. Hence it is that the 
ordinary gas scales and the thermodynamic 
scale are all approximately the same, and the 
problem of finding the mutual relations of the 
various scales is reduced to the investigation 
of the departures of the actual gases from the 
ideal state and the computation of corrections 
for-the departures."* 

The kinetic theory of gases enables us 
probably to obtain a clearer conception of 
what is actually meant by an "ideal gas." 
According to this theory trie pressure exerted 
by any gas against the walls of the containing 
vessel is due to bombardment by a large 
number of infinitesimally small particles 
(molecules) which are in rapid motion. The 
pressure therefore increases with the number 
of molecules per unit volume and their aver- 
age velocity. The volume actually occupied 
by the molecules themselves is assumed to be 
infinitesimal as compared with the volume 
in which they are present. The collisions 
between the molecules must be perfectly 
elastic, otherwise the energy of the gas would 
tend to decrease indefinitely. Heat applied 
to the gas is converted into kinetic energy 
of the molecules; thus the average kinetic 
energy forms a measure of the temperature 
of the gas. 

When the gas expands, heat is absorbed 
because the molecules have to perform a 
certain amount of work against the external 

*E. Buckingham, Bull. Bur. Standards, 3, 237 (1907). 
Reprint No. 57. 

pressure acting on the walls. In a perfect 
gas the amount of heat absorbed is exactly 
equal to the amount of external work done. 
If, however, additional energy is required 
to overcome any attractive forces between 
the molecules themselves, the amount of 
external work will be less than the total 
energy absorbed. Similarly, if the forces 
between the molecules are repulsive, the dis- 
crepancy between external work and heat 
absorbed is in the opposite direction. We are 
thus led to conceive of a perfect gas as one 
in which the volumes of the molecules are 
practically reduced to points, while the forces 
acting between them are diminished to a 
negligible factor. 

From the kinetic point of view the absolute 
zero is the temperature at which the molecules 
of a gas lose all kinetic energy. That such a 
state may be impossible to realize in practice, 
only leads to the further belief that we can 
never actually attain the absolute zero. 

Gay-Lussac's Experiment 

The notion of a perfect gas arose from 
two facts: first, the validity of Boyle's law 
over very large ranges of pressures and tem- 
peratures for nearly all the ordinary gases, 
and second, the demonstration by Gay- 
Lussac that the temperature of a gas does not 
change by any noteworthy amount when the 
volume merely increases without doing ex- 
ternal work. 

The latter experiment has become a classic 
in the history of science. Gay-Lussac con- 
nected two receivers by means of a tube 
furnished with a stop-cock, and immersed 
them in a bath of water which served as a 
calorimeter. One of the receivers was filled 
with air under pressure, the other was ex- 
hausted. On opening the stop cock between 
the two vessels, the pressure naturally be- 
came equalized. However, the temperature 
of the surrounding water remained at the 
same value as before the expansion. The 
conclusion drawn was therefore that "no 
change of temperature occurs when air is 
allowed to expand in such a manner as not to 
develop mechanical power, "f 

Porous-Plug Experiment 

The experiment was subsequently repeated 
by Joule and Thomson under much more 
accurate conditions and the conclusion of 
Gay-Lussac shown to be not quite true for 
ordinary gases. The latter investigation is 
known as the "porous-plug" experiment 
and for a detailed description of this the 

t Preston, p. 286. 



reader may refer to any text-book on heat. 
In this experiment the gas is forced to flow 
steadily through a porous plug, which is so 
insulated that no heat can enter or leave it by 
conduction. The pressure and temperature 
are observed on both sides of the plug and 
from these observations it is possible to 
determine whether there is any change of tem- 
perature when the gas expands without per- 
forming mechanical work. It was found by 
Joule and Lord Kelvin that hydrogen becomes 
warmer, while all the other ordinary gases 
become colder in passing through the capil- 
laries of the plug. 

With the data obtained from the porous- 
plug experiment and the further observations 
on the manner in which the different gases 
deviate from Boyle's law at different pressures 
and temperatures, it is therefore possible 
to calibrate the indications of the ordinary 
air or hydrogen gas thermometer (the usual 
form is that at constant pressure) in terms of 
absolute or thermodynamic scale.* 
Radiation Scale of Temperature 

It was shown experimentally by Stefan, 
and deduced theoretically by Boltzmann that 
the radiation within an enclosure whose walls 
are maintained at a uniform temperature 
is absolutely independent of the nature of the 
enclosure, and varies with the fourth power 
of the temperature. If E denotes the amount 
of energy radiated per unit area of a black- 
body radiator at temperature 0, we have the 

E = bd* 

At the absolute zero, the energy radiated 
is zero, and if we choose a suitable value of 
the constant b, we. can make the radiation 
scale of temperature agree with the absolute 
scale at all temperatures. We are thus pro- 
vided with another method of calibrating 
thermometers and pyrometers in terms of the 
thermodynamic scale. 
Methods of Attaining Low Temperatures 

The different methods which have been 
used for attaining low temperatures may be 
classified under the following heads: 

(1) Methods involving the use of freezing 

(2) Methods involving the liquefaction of 
gases under pressure and the subsequent evap- 
oration of these liquids. 

(3) Cooling of gases owing to adiabatic 

(4) Cooling of gases owing to the Joule- 
Thomson effect. 

Freezing Mixtures 

The addition of salt to water lowers its 
freezing point by an amount which increases 
with the concentration of the salt in solution. 
On the other hand, the solubility of salt in 
water decreases with the temperature. Con- 
sequently, at a certain definite temperature 
the solution freezes as a whole, the composi- 
tion of the solution being the same as that of 
the ice-salt mixture which separates out. 
This temperature, which is the lowest at which 
salt and ice can exist in equilibrium with a 
solution of salt in water, is known as the 
cryohydric temperature, and the mixture of 
ice and salt which has this definite melting 
point is known as a cryohydrate. From these 
facts it is easy to give an explanation for the 
cooling effect of such freezing mixture. 

If we mix ice, salt and water at a tempera- 
ture above the cryohydric point, the water will 
tend to dissolve salt until it becomes saturated; 
on the other hand, ice will melt and tend to 
dilute the solution which will again dissolve 
more salt, and this will result in the melting 
of more ice. As the freezing mixture is 
assumed to be well insulated thermally, the 
temperature must decrease owing to the 
latent heat abstracted for melting the ice, 
until finally a temperature is attained at 
which the solution has the same composition 
as the ice and salt mixture which separates 
out from it on freezing. 

The temperatures obtainable by the use of 
cryohydrates range as low as —55 deg. C. 
The following table gives the compositions 
of a few of these cryohydrates together with 
the temperature of the cryohydric point. 



Percentage of 

Name of Salt 


Anhydrous Salt 
in Freezing 



Calcium chloride 



Sodium bromide 



Sodium chloride 



Sodium nitrate 



Ammonium chloride . . 



Magnesium sulphate. 



* E. Buckingham, Bull. 57, Bur. of Standards. 

Liquefaction of Gases by Pressure 

The freezing of water at ordinary tempera- 
tures owing to very rapid evaporation is a 
familiar phenomenon. The temperature of 
any liquid tends to maintain itself at that 
value which corresponds to the pressure of 
the vapor above it. If now a vessel of water 
is placed under the receiver of an air-pump 



and the water vapor pumped out very rapidly, 
the temperature of the water is decreased, 
owing to the heat absorbed in evaporation, to 
a point at which the pressure of the water 
vapor in the receiver is in equilibrium with 
the water. 

Similarly the temperature of any liquid can 
be lowered very considerably if it be allowed 
to evaporate very rapidly, and upon this 
principle depends the use of liquefied gases 
in producing low temperatures. 

It is possible, by the use of very high pres- 
sures, to liquefy a large number of gases at 

Atm —- 
Fig. 1. Decrease in.Temperature of Air Obtainable with Adiabatic 
Expansion (full line curves) and Joule-Thomson 
Expansion (dotted lines) 

ordinary temperatures. Faraday was one of 
the first experimenters to make extensive 
use of this method for liquefying gases. 
He used thick-walled glass U-tubes, generated 
the gas in one limb and allowed it to condense 
under its own pressure in the other limb, 
which was cooled in an ice-salt mixture. The 
following gases were condensed by him by 
this method: S0 2 , HL, CL 2 , NH 3 , C 2 N 2 , 
H 2 S, HBr, PH 3 , HC1, N 2 0, C0 2 and C 2 H 4 . 

By evaporating liquid C0 2 at ordinary pres- 
sure it was found possible to attain a temper- 
ature of —78 deg. C. In present practice, 
the liquid carbon dioxide contained in a steel 
cylinder is allowed to evaporate at atmospheric 
pressure; and owing to the rapid evaporation 
some of the out -flowing liquid solidifies to a 
snow-like mass. This is mixed with ether or 
toluene and used for maintaining a constant 
temperature of —78 deg. C. 

In the case of liquefied ammonia and sul- 
phur dioxide, the temperatures obtainable by 
allowing these to evaporate at atmospheric 
pressure are —31 deg. C. and —10 deg. C. 
respectively. By allowing these liquids to 
evaporate at 0.1 atmospheric pressure, the 
temperature is lowered still more, and in the 
case of C0 2 , Faraday obtained a temperature 
of —110 deg. C. by evaporating the liquid 
under very low pressure. 

Critical Temperature 

There remained some gases, however, which 
Faraday and subsequent experimenters were 
unable to condense even with the highest 
available pressures. These were therefore 
designated as "permanent" gases, and 
included CH 4 , NO, C0 2 , 2 , N 2 and H 2 . 
Subsequently there were added to this list 
the so-called' "noble" or rare gases, krypton, 
argon, neon and helium. 

Andrews (1869) first pointed out that ex- 
tremely great pressure alone is not sufficient 
for the liquefaction of gases. It is also neces- 
sary to cool the gas below its critical tempera- 
ture. At this temperature the gas may be 
condensed by a pressure which is known as 
the critical pressure, while at higher tempera- 
tures the gas remains incondensable no matter 
how high the pressure is raised. The following 
table gives the critical temperature and criti- 
cal pressure of a number of gases.* 





Carbon monoxide. . . 



Nitric oxide 


Carbon dioxide 

Hydrogen chloride . . 

Ammonia i NH 

Sulphur dioxide SOj 

Water I H»0 

Mercury I Hg 


H 2 

N 2 















- 92.9 

- 81.8 

















* K.'Jellinek. Lehrbuck der physikal. Chemie. I (1), p. 444. 



It can be observed from this table that the 
gases which had been condensed by Faraday 
and others before the investigations of 
Andrews have this feature in common, that 
their critical temperatures lie above deg. C. 

As a result of Andrews' work, it became 
evident that in order to liquefy the so-called 
permanent gases it is necessary to cool them 
to temperatures still lower than those hitherto 

Cooling of Gases by Adiabatic Expansion 

When a gas is allowed to expand reversibly, 
that is, in such a manner that the pressure 
of the gas is always just equal to the external 
pressure, work is done against this external 
pressure. If the gas is maintained at con- 
stant temperature, the expansion is said to be 
isothermal, and the energy required to over- 
come the external pressure is absorbed from 
the constant temperature reservoir. If, 
however, the gas is insulated thermally, so 
that heat can neither enter nor leave it, the 
energy required for expansion is absorbed 
from the kinetic energy of the gas molecules 
themselves, so that the temperature of the 
gas decreases. Such an expansion is said to 
be adiabatic, and the relation between the 
pressure and temperature in the case of an 
ideal gas is given by the equation 

^ . .. JC-l 



where A" is a constant for each gas. In the 
case of air the value of this constant is 1.40. 

The curves shown in Fig. 1 indicate the 
cooling effect to be expected by expanding 
air adiabatically at different initial tempera- 
tures from higher pressures to 1 atmosphere. 
Thus starting with air at 20 deg. C. and 50 
atmospheres, the temperature can theoreti- 
cally decrease to -177 deg. C. (96 deg. K.)* 
If the compressed gas is initially cooled to 
— 60 deg. C, the lower limit of temperature 
becomes -204 deg. C. (69 deg. K.). It is 
evident therefore that it is possible to obtain 
considerable cooling bv adiabatic expansion 

L. Cailletet condensed in this manner the 
gases oxygen, nitrogen and carbon monoxide 
in small amounts. More recently Claude has 
applied the same principle to the construction 
of an apparatus for the continuous production 
of liquid air. 

Cascade Method of Liquefying Gases 

R. Pictet developed a very useful method 
of liquefying gases which has since then been 

* We will use the symbol "deg. K." to denote degrees abso- 
lute (Kelvin scale). 

applied to great advantage by Kammerlingh 
Onnes in his cryogenic laboratory at Ley den. 
The fundamental principle of this method 
consists in cooling a gas (A) below its critical 
temperature by the rapid evaporation of 
another condensed gas, then using the lique- 
fied gas A to cool a third gas B below its 
critical temperature, and so on. 

Kammerlingh Onnes attains a constant 
temperature of —217 deg. C. (56 deg. K.) 
as follows: Methyl chloride is condensed by 
pressure at ordinary temperature and then 
allowed to evaporate under reduced pressure. 
This produces a temperature of —90 deg. C. 
and is used to condense ethylene (critical 
temperature, 10 deg. C). The latter, in turn, 
when evaporated under reduced pressure 
produces a temperature of —165 deg., which 
is below the critical temperature of oxygen. 
By liquefying the last gas at the temperature 
of — 165 deg. and evaporating the condensed 
product under reduced pressure it is possible 
to obtain a temperature of —217 deg. and 
maintain it for quite a long time. 

Fig. 2 illustrates the method diagram- 
maticallv. The methvl chloride is condensed 

/ \ 




Fig. 2. Cascade Method of Liquefaction 

by the compressor B and drawn through d 
into the outer tube of the condenser C. It 
is there evaporated by the exhaust pump A 
at a temperature of —90 deg. C. The 
ethylene passes through a similar cycle by 
means of the compressor F and exhaust pump 



E, producing a temperature of —165 deg. 
in H. The oxygen is generated in L and con- 
denses under its own pressure in the tube M. 

A consideration of the vapor pressure 
curves of different gases as drawn in Fig. 3* 
shows that this method is not applicable to 
the liquefaction of either hydrogen or helium, 
since the critical temperatures of both these 
gases are much below the lowest temperatures 
obtainable by evaporating the gases of higher 
boiling point. 
Cooling of Gases Owing to Joule-Thomson Effect 

If a gas is allowed to pass through a capil- 
lary tube from a higher to lower pressure, it 
ought, in the ideal case, to show no change 

• Normal Boiling Point 
o Critical Point. 

-213 SO tO W 200 SO SO *0 20 -100 SO SO HI 20 20 40 SO SO f IOO 

Deqrees -» 

Fig. 3. Liquefaction Temperatures of Gases at 
Different Pressures 

in temperature, since no external work is 
gained or lost. The porous plug experiment 
which has already been mentioned, shows 
however, that all ordinary gases, with the 
exception of hydrogen and helium, experience 

* K. Jellinek, loc. cit. p. 451. 

a lowering of temperature when expanded 
through a capillary. In the case of hydrogen 
and helium, there is a heating effect; but this 
gradually diminishes as the temperature at 
which expansion occurs is lowered, and below 
the so-called inversion temperature ( — 90 

Fig. 4. Linde's Process for Liquefying Air 

deg. C. for hydrogen) there is a cooling effect 
as in the case of the other gases. 

Linde's process for the liquefaction of air 
depends upon this principle. It is illustrated 
diagrammatically in Fig. 4. The air is com- 
pressed to 20 atmospheres in the compressor 
e and then passes into the smaller compressor 
d where the pressure is increased to 200 atmos- 
pheres. The gas then passes through the 
pipe Pi and the water-separator / into the 
cooling spiral g which is immersed in ice and 
salt (-20 deg. C). The cooled gas then 
passes through the series of pipes P 2 and 
expands in the nozzle o to 20 atmospheres. 
The temperature falls during this operation 
to —78 deg. C. By passing this cooled air 
over other pipes carrying air at an initial 
temperature of —20 deg. C, the latter is 
cooled still further, so that when it expands 
in the nozzle a, its temperature falls well 
below —78 deg. By using three systems of 
concentric pipes the air is finally cooled down 
to a temperature at which it can be liquefied. 
The expansion valve b is then opened, so that 
a fraction of this cooled air expands from 20 
atmospheres to 1 atmosphere and condenses. 
The liquid air is collected in the Dewar 
flaskf C. 

t In order to prevent the rapid evaporation of liquid air and 
similar products, Dewar suggested the use of double-walled 
flasks in which the space between the walls has been well evacu- 
ated. A vacuum is the best heat insulator known. 



The cooling effect actually produced in a 
Linde machine under operating conditions 
is shown in Fig. 4 by the fine dashed lines. 
Thus, air at — 20 deg. is cooled to — 36 deg. by 
expanding from 50 atmospheres to 1 atmos- 
phere. If this cooled air is compressed and 
again expanded, the temperature drops to — 54 
deg., and then to —101 deg., —136 deg. and 
finally —190 deg. At this temperature liquid 
air has a vapor pressure of 1 atmosphere, 
so that the expanding air condenses. 

In 189S Dewar succeeded in liquefying 
hydrogen by the same method. As this gas 
has an inversion temperature of —90 deg. 
C, he cooled it in liquid air before subjecting 
it to the Joule-Thomson process. 

By evaporating liquid hydrogen at low- 
pressure it becomes possible to obtain tem- 
peratures ranging from —252 deg. C. (the 
boiling point at atmospheric pressure) to 
-259 deg. C, that is, from 21 deg. K. to 
14 deg. K. This is still, however, above the 
critical temperature of helium. 

As the inversion temperature of this gas 
occurs at 33 deg. K., its liquefaction presented 

immense difficulties. In 190S, Kammerlingh 
Onnes succeeded in liquefying helium by 
cooling the gas first in liquid hydrogen and 
then cooling it still further by the Joule- 
Thomson process. The boiling point of 
helium at atmospheric pressure is 4.29 deg. 
K., and by evaporating the liquid under 
reduced pressure, Onnes has been able to 
obtain a temperature of 1.48 deg. K. ( — 271.6 
deg. C). These temperatures were measured 
by means of a low pressure helium ther- 
mometer, on the assumption, of course, that 
the gas laws are perfectly valid for helium 
under these conditions. Whether this 
assumption is justifiable cannot as yet be 
definitely stated. This much, however, is 
certain, that by evaporating liquid helium 
under very low pressure we are able to obtain 
a temperature which is within less than two 
degrees of the absolute zero. 

In the next issue we shall discuss the 
behavior of different substances at these low 
temperatures, and point out the theoretical 
importance of the study of these low-tem- 
perature phenomena. 


By C. W. Larson 
Industrial Locomotive Designing Engineer, General Electric Company 

The January, 1914, number of the General Electric Review contained eight articles describing the 
electrical and mechanical controlling devices for the lock machinery of the Panama Canal; and the July, 1914, 
number contained three articles describing the hydraulic turbines and equipment, the electric generators, and 
the controlling switchboard equipment located in the power station at Gatun which furnishes energy for operat- 
ing the canal. The following article describes the ship towing locomotives used at the locks. The first part of 
the article is devoted to the presentation of the reasons why none of the hitherto existing systems of maneuver- 
ing ships in close quarters could be satisfactorily applied to the locks of this canal. Following this is a descrip- 
tion of the system developed to fulfill the conditions. Next is a minute description of the locomotives them- 
selves. These, it is very satisfying to know, are fully in keeping with the other unique devices which have 
been developed to make this wonderful canal possible. — Editor. 

The President of the United States in June, 
1905, appointed a Board of Consulting 
Engineers, men of international reputation, 
"for the purpose of considering the various 
plans proposed to and by the Isthmian Canal 
Commission for the construction of a canal 
across the Isthmus of Panama." 

The majority report of this Board of Con- 
sulting Engineers contains the following: 

"The three accidents at St. Mary's Falls 
Canal occurred within a period of nine years, 
where there is only one lockage of about 20 
feet. If six locks should be adopted in a 
plan for the Panama Canal, each having a 
lift of 30 feet or more, as has been proposed 
in several projects, it would not be unreason- 
able, with an equal number of vessels, to look 
for six times the number of accidents in the 
same period of time, which would be at the 
rate of two per year. If groups of locks 
should be arranged in flights, as has also been 
proposed in some projects, the imminence of 
disastrous accidents would be greatly en- 
hanced, as would be the amount of damage 
to the structures and to the vessels involved. 
Indeed, it is highly probable that the grave 
disaster of a great ocean steamship breaking 
through the gates of the upper lock and 
plunging down through those below might be 

It is true that the majority of the Board 
favored the adoption of the sea-level canal, 
but the foregoing quotation showed the 
necessity, in their opinion, of safeguarding 
the passage of vessels through the locks. 

Investigations of collisions between ships 
and lock gates invariably show that "there 
was a misunderstanding in signals between 
the captain and the engineer." Bearing in 
mind that the engineer of the ship is so 
situated that he does not know the exact 
position of the ship, with respect to the lock, 

he cannot check his actions by the probable 

A system, therefore, which permits the 
checking of the movement of the ship with 
the signal given by the pilot or captain of 
the ship will eliminate improper manipula- 
tions to a very great extent. 

Various systems are in vogue at dry-docks 
which are based on the principle that the 
operator sees the result of his action. The 
employment of winches or capstans has been 
looked upon with a great deal of favor. 
These are usually placed at intervals along 
the dock walls, and the lines from the ship 
are carried forward to the successive capstans 
as the ship advances. Such a system in- 
volves the risk of the ship not being properly 
safeguarded when the lines are transferred 
to the successive capstans. An improve- 
ment has been made by the installation of a 
capstan at the head of the dock, centrally 
located, and used for imparting a straight 
motion to the ship. Numerous lines from the 
ship to the dock wall are carried by men, and 
the capstans are employed to counteract 
any wind pressure, currents, etc., and assist 
generally in maneuvering the ship. While 
an improvement over former methods, it 
did not, however, possess the flexibility and 
reliability required for the operation of the 
locks of the Panama Canal, neither would 
it have eliminated the breaking of the lines 
at critical moments, which is regarded as one 
of the essential requirements in successfully 
handling ships in canal locks. 

After a very thorough study of the entire 
problem of maneuvering the ships through 
the locks of the Panama Canal, it became 
evident that the ships should not proceed 
through the locks under their own power, 
and that a substitute for the ship's power 
should embrace the following requirements : 



(a) The ability to place the ship in 

proper relation to the lock. 

(b) The capability of keeping the ship 

in its course. 

(c) The accelerating and retarding of 

the ship without rupturing the 

(d) The lines when once attached should 

be used without change for lock- 
age in flight. 

(e) The services of a small number of 

skilled operators rather than a 
large number of unskilled men. 
The towing system described in the fol- 
lowing pages was designed and patented by 
Mr Edward Schildhauer, Electrical and 
Mechanical Engineer of the Isthmian Canal 

Towing System 

In passing through the canal from the 
Atlantic to the Pacific, a vessel will enter the 
approach channel in Limon Bay, which ex- 
tends to Gatun a distance of about seven 
miles. At Gatun it will enter a series of three 
locks in flight and be raised 85 feet to the 
level of Gatun Lake. It may then steam at 
full speed through the channel in this lake, 
for a distance of 24 miles, to Bas Obispo, 
w-here it will enter the Culebra Cut. It will 
pass through this cut, which has a length of 
nine miles, and reach Pedro Miguel, where 
it will enter a lock and be lowered 30 feet. 
Then it will pass through Miraflores Lake for 
a distance of one and one-half miles until 
it reaches Miraflores, where it will be lowered 
55 feet through two locks, to the sea level, 
after which it will pass into the Pacific 
through an eight and one-half mile channel. 

The main features of all the lock sites are 
identical and the following brief description 
of the Gatun Locks, with especial reference 
to the arrangement of the towing tracks, 
ship channels, inclines and approaches, is 
given to present a clearer conception of the 
towing scheme in general. A more detailed 
description of the locomotive itself will then 

The general layout of the Gatun Locks is 
clearly shown in Fig. 1. It will be noted that 
there are two ship channels, one for traffic 
in each direction. The channels are separated 
by a center wall, the total length of which is 
6330 feet. There are two systems of tracks 
for the locomotives, one which they use when 
towing and the other when they are returning 
idle. This, however, refers onlv to the outer 

walls, since for the center wall there is only 
one return track in common for both the 
towing tracks. The towing tracks are 
naturally placed next to the channel side, 
and the system of towing normally utilizes 
not less than four locomotives running 
along the lock walls. Two of them are op- 
posite each other in advance of the vessel, 
and two run opposite each other following 
the vessel, as seen in Fig. 2. The number 
of locomotives is increased, however, when 
demanded by the tonnage of the ship. 

Cables extend from the forward locomo- 
tives and connect respectively with the port 
and starboard sides of the vessel near the 
bow, and other cables connect the rear 
locomotives with the port and starboard 
quarters of the vessel. The lengths of the 
various cables are adjusted by a special 
winding drum on the locomotive so that the 
vessel will be placed substantially in mid- 
channel. When the leading locomotives are 
started they will tow the vessel, while the 
trailing locomotives will follow and keep all 
the cables taut. By changing the lengths 
of the rear cables the vessel can be guided, 
and to stop it, all the locomotives are slowed 
down and stopped, thus bringing the rear 
locomotives into action to retard the ship. 
Therefore, the vessel is always under com- 
plete control thoroughly independent of its 
own power, and the danger of injury to the 
lock walls and gates is thereby greatly les- 

The illustration in Fig. 12 shows how effec- 
tively the four locomotives keep the vessel 
under control and in the center of the channel, 
while Figs. 8 to 13 give a general idea of the 
method of handling vessels of various sizes. 
They also show general views of the lock 
walls, towing tracks, and inclines, the steep- 
ness of the latter being especially noticeable. 
Of particular interest is Fig. 11, which repre- 
sents a trial tow approaching the second 
level. The water in the middle lock or at 
this second level is at sea-level, a condition 
not obtained in regular operation; and this 
trial was made to demonstrate that the tow- 
lines would clear the lock walls. 

The towing tracks have a specially de- 
signed rack-rail extending the entire length 
of the track and centrally located with 
respect to the running rails. It is through 
this rack-rail that the locomotive exerts 
the traction necessary for propelling large 
ships and for climbing the steep inclines. 

Rack-rail is also provided on short portions 
of the return track so as to lower the loco- 



Fig. 1. A Drawing of the General Layout of the Gatun Locks. The heavy dot-dosh line indicates the portion of the towing locomotive track that is equipped with rack-rail 

Pig. 2. Diagram Showing the Relative Location of the Locomotives, Cables and Ship During 
the Operation of Towing 


Fig. 3- A Section of the Fixed Rack-Rail at the Left to 

which a Section of Movable Rack-Rail is Hinged at C. 

One of these movable sections is located at each 

end of the rack-rail 

Fig. 4. A Section of the Conduit Showing the Current Collecting 

Device for Two Phases and the Underground 

Rails from which the Current is Taken 


Fig 6 The Plan View of a Towing Locomotive with the Covers Removed from the left-hand End 

Fig. 7. A Cross Sectional View. Near tbe Winding I 
of a Towing Locomotive 



motives safely from one level to the next. 
The steepest slope is 26 deg. or 44 per cent, 
hence the need will be seen for rack-rail even 
on the return track, it being 
known that any traction loco- • — 
motive with the usual wheel 
drive, even with brakes set, 
would begin to slide on a 10 
per cent grade and could there- 
fore not be controlled. With 
a rack-rail, however, traction 
is limited only by the capacity 
of the driving motors and not by 
the adhesion of the wheel treads 
to the rails. 

A small portion of rack-rail 
is shown in Fig. 3, A being the 
rack-rail proper and B the 
approach to it. B is hinged at 
C so that it can be depressed on 
the approach of the rack-pinion 
of the locomotive. The teeth 
of the approach section are 
under size and are shaped off at 
the extreme end so that the 
teeth of the pinion will mesh 
properly and thus prevent ex- 
cessive strain on the pinion and 
the axle. The spring D restores 
the approach to proper position after the 
locomotive has passed over. The rack-rail is 
of the shrouded type, and each tooth space 

the walls. A further feature of the rack-rail 
is the projecting edges which permit thrust- 
wheels attached to the locomotive to run 

Fig. 9. The "Ancon" Entering the Upper Gatun Lock from the Middle West 
Chamber under the Tow of Electric Locomotives 

has a drain-hole cast in the bottom to carry 
off water and other accumulations to suitable 
drain pipes or ducts set in the concrete of 

Fig. 8. Electric Locomotives Locking 85-ft. Piles through the Pedro-Miguel Locks 

along the under side and prevent the over- 
turning of the locomotive, in case some un- 
foreseen operating condition should produce 
an excessive pull on the tow- 
line. These thrust-wheels serve 
to counteract the lateral com- 
ponent of the tow-line pull 
and the flanges act for emer- 
gency only, as the weight of the 
locomotive is sufficient to pre- 
vent overturning with the nor- 
mal pull of 25,000 pounds on the 
tow-line. These thrust-wheels 
are shown in Fig. 7. 

Three-phase, 25-c)rcle, 220- 
volt alternating current is used 
for operating the locomotives, 
and the current is supplied to 
the locomotives through an 
underground contact system . 
The collecting device is illus- 
trated in Fig. 4, while Fig. 7 
shows its position with respect 
to the track, it being adjacent 
to the running rail on the side 
remote from the lock. Two 
T-rails (shown in black section) form two 
legs of the three-phase circuit and the third 
leg is formed by the main track rails. A 



Fig. 10. 

The First Trial Run of the Towing Locomotives at the Gatun Locks. 
in making this test 

A barge was used 

specially designed contact plow slides between 
the two T-conductors and transmits the 
power from the rails to the locomotive. 
This contact plow passes through the slot 
opening in the conduit cover and is flexibly 
connected to the locomotive in such a manner 
as to follow all irregularities in the tracks and 
crossovers, and therefore insures a continuous 
supply of power. 

Locomotive Design 

The working parts of the locomotive are 
supported by two longitudinal upright cast- 
steel side frames, No. 1, Fig. 5, connected by 
transverse beams, No. 2, Fig. 6. These 
frames are, in effect, deep rigid trusses, hav- 
ing upper and lower members connected by 
posts, No. 5, and diagonal braces, No. 6, 
Fig. 5. The middle portion of each frame 

Fig. 11. The First Trail Tow with the Barge at the Second Level of the Gatun Locks, Rear Locomotives 

Ascending the Incline. The water in this middle lock (second level) is shown at sea level, 

a condition not obtained in regular operation. This trial was made to 

demonstrate that the tow lines would clear the lock walls 



Fig. 12. Four Towing Locomotives Attached to the Submarine Tender "Severn" in a Gatun 

Lock Ready for Lowering the Water-Level. A group of submarines may 

be seen at the far end of the lock 

has its upper and lower members parallel 
and horizontal, but the end portions have 
their lower members inclined upward toward 
the ends of the frame. The pedestals, No. 
7, for the wheel axles, No. 8, are located at 
the junction of these end portions with the 
middle portion, and are of the usual locomo- 
tive type, having vertical parallel jaws 
between which the journal, No. 9, slides. 
Springs, No. 10, are interposed between the 

tops of the journal boxes and the tops of the 
pedestals, and the locomotive is thus mounted 
upon four wheels, No. 11, carried on the two 
axles, No. 8, the wheel-base being 12 feet and 
the overall length of the locomotive over 32 

Each axle is driven by its own motor, in- 
dependent of the other, and as the construc- 
tion is identical at both ends of the machine, 
a description of one end will suffice for both. 

Fig. 13. The "Severn" Leaving the Upper East Chamber in the Tow of Electric Locomotives 



A cast-steel suspension bracket, No. 12, Fig. 
14, is hinged at one end upon the axle. Its 
bearings, No. 13, which fit the journals on 
the axle are secured in place by caps, No. 14, 

Fig. 14. A Side Elevation of the Traction Brackets 

which are provided with oil cellars. No. 15. 
The bracket is provided with bearings for a 
transverse jackshaft, No. 16, parallel with 
the axle, and it has pillow blocks. No. 17, for 
a countershaft, No. 18, also parallel with the 
axle. It has a substantial horizontal plat- 
form, No. 19, to support the driving motor, 
No. 20, and its outer end is supported at each 
corner by two springs, No. 21, placed above 
and below a stationary angle-iron, No. 22, 
and connected to the bracket by a bolt. No. 
23, so as to afford a yielding support in both 
upward and downward movements of the 

3940 I Ui^>C — ~-& 

Fig. 15. A Longitudinal Horizontal Section of 
a Clutch Shaft 

The motor, No. 20, is of the three-phase, 
slip-ring type, enclosed, and identical to the 
rugged steel-mill design, and it is geared 
by pinion, No. 24, and spur gear, No. 25, 

to the countershaft, No. 18, which carries 
a pinion, No. 26, meshing with a spur gear, 
No. 27, Fig. 15, keyed to the jackshaft, No. 

16. On the outer side of the spur gear, No. 
27. are formed clutch teeth which cooperate 
with similar teeth, No. 28, on the adjacent 
side of a gear, No. 29, which is sleeved upon 
the jackshaft, and which can be slid length- 
wise thereon to engage and disengage the 
clutch teeth. The means for sliding this gear 
consists of a disk, No. 30, secured to the gear 
and having a central hub, No. 31, fitting over 
the end of the jackshaft, Fig. 15. A rod, No. 

32, Fig. 15, runs through a central hole in 
the shaft and through the center of the hub. 
No. 31, to which it is connected by nuts, No. 

33, in such a manner as to permit the disk 
to rotate with the wheel, and at the same time 
to cause it to slide the wheel axially when the 
rod is reciprocated. A pinion, No. 34, Fig. 

17, is keyed to the axle, No. 8, and is wide 
enough to always mesh with the gear, No. 
29, so that when the clutch teeth, No. 28, 
are engaged, the motor will propel the loco- 
motive by the adhesion between the wheels. 
No. 11, and the rails of the track, and this 
only when running without load and between 

When the locomotive, however, reaches 
one of the inclines between the locks, the 
grade of which may be as much as 44 per cent, 
or when it is towing a ship, the cog-rail 
system is utilized to enable the locomotive 
to climb the grade or to exert the traction 
necessary for pulling large ships. The cog 
or rack-rail is laid between the track rails, 
and the locomotive is provided with a cog 
wheel or rack pinion, No. 35, Fig. 17, secured 
to or integral with a sleeve, No. 36, which 
rotates freely on the axle. A gear wheel, No. 
37, secured to or integral with this sleeve, 
meshes with a gear, No. 38, Fig. 15, which 
turns loosely on the jackshaft. Clutch teeth, 
No. 39, on this gear can be engaged by teeth, 
No. 40, on a clutch, No. 41, which is splined to 
a jackshaft. A two-armed lever, No. 42, 
fulcrumed on a bracket, No. 43, straddles 
the shaft, No. 16, and is pivoted to a collar, 
No. 44, riding in a groove in the clutch, No. 
41. The lever is connected by a link, No. 45, 
to one end of a lever, No. 46, Fig. 15, turning 
loosely on a vertical rockshaft, No. 47. An 
elastic arm is keyed to 'the shaft and engages 
lugs on the lever, No. 46. The arm, No. 4S, 
is composed of a laminated flat-steel soring, 
Fig. 16, and a second arm, No. 49, is keyed 
on the shaft and connected by a rod, No. 50, 
to a handle in the cab of the locomotive. The 



handle can be locked by a suitable latch and 
notched quadrant, and the other end of the 
lever, No. 46, is pivoted to the rod, No. 32, 
so as to throw out the clutch, No. 28, when the 
clutch, No. 40, is thrown in, and vice versa. 

The elastic arm, No. 48, serves to throw 
the clutches automatically; it being under- 
stood that the four-jaw clutches in most 
cases do not mesh when thrown but that the 
operating handle is thrown full stroke and 
locked. This puts the springs under heavy 
tension. The locomotive is then started 
slowly and when the clutches are in align- 
ment the springs throw them without any 
attention by the operator. 

The two rocker shafts at opposite ends of 
the locomotive are connected by the rods, 
No. 52, pivoted to rocking arms, No. 52a, on 
the shafts and to an intermediate lever, 
No. 52b, fulcrumed on the pedestal sup- 
porting the winding drum. Figs. 16 and 19 
will be of assistance in making clear the 
foregoing description of clutches and gear. 

Fig. 16. A Drawing of the Clutch Operating Mechanism 

The two traction motors, No. 20, are con- 
trolled by suitable controllers installed in the 
cabs at the ends of the locomotives, and the 
circuits are such that both motors can be 
controlled from either cab, and can be 
operated singly or in multiple as desired. 
Current is taken from the supply conductors 
by the special current-collecting device pre- 
viously described and shown in Figs. 4 and 

It will be observed that each motor, with 
all its gearing and clutches, is mounted in- 
dependently of the frame of the locomotive, 
to which it is connected only by the springs, 
No. 21, which give an elastic support for the 
outer end of the bracket, No. 12, Fig. 14, on 
which the mechanism is carried. 

In connection with each motor a powerful 
brake is installed, and as during operation the 
motors are at all times geared either to the 
axles or to the cog wheels, the truck wheels, 
No. 11, are not provided with any brake 
rigging. The motor brake is shown in Figs. 
5 and 6, but it is more clearly illustrated 

in Figs. 18 and 20. On the motor shaft is 
keyed a brake disk or drum, No. 53, Fig. 18, 
and to opposite sides of it are applied the 
brake shoes, No. 54, carried bv the brake 

ij nrr 

Fig. 17. 



? c 


D C 

A Longitudinal Horizontal Section of 
an Axle Shaft 

levers, No. 55, which are pivoted at No. 56 
upon a stationary bar, No. 57, projecting 
from a frame, No. 58, which supports a 
solenoid, No. 59. The movable core of this 
solenoid is pivoted to the long arm of a lever, 
No. 60, which is fulcrumed at No. 61 on one 
of the brake levers. A rod, No. 62, connects 
the angle of this lever with the other brake 
lever, thus constituting a sort of toggle 
between the two levers. When the core of 
the solenoid drops, it actuates the lever and 
the rod in such a manner as to draw the two 
brake levers toward each other, thereby 
applying the brake shoes to the drum. The 
winding of the solenoid is in circuit with the 
controller of the motors, so that when the 

Fig. 18-5 A Side Elevation of the Combination Hand and 
Solenoid Brake and Rigging 

current is turned on to energize the motor 
windings, the solenoid will lift its core and 
thereby release the brakes. The first point 
of the controller releases the brakes without 
applying power to the motors, thereby pro- 



viding a coasting point. But should the motor 
current be shut off, either intentionally or 
accidentally, the core will instantly drop by 
gravity and its weight will exert a powerful 

Fig. 19. A Front View of a Traction Motor Unit with 
a Journal Box in Place 

leverage upon the brake levers to stop the 
motors and the locomotive. This action 
occurs simultaneously on both motors, and 
brake action is powerful enough to stop the 
locomotive within two revolutions of the 

In addition to this automatic brake, means 
are provided for applying the brakes manually 
in order to supplement the action of the auto- 
matic feature, 'if necessary, when descending 
a grade or when approaching a rack-rail. An 
upright shaft, No. 63. Fig. IS, provided with 
a hand-wheel, No. 64, has attached to it one 
end of a chain, No. 65, which runs under a 
stationary pulley, No. 66, up over another 
pulley. No. 67, on one end of an elbow-lever. 
No. 6S, pivoted to one of the brake levers, 
and thence under a stationary pulley, No. 69, 
to the opposite end of the locomotive. The 
elbow-lever, No. 6S, has its other arm con- 
nected by a rod. No. 70, to the other brake 
lever, the rod being adjustable in length as 
shown. The lever. No. 6S, and rod, No. 70, 
constitute a toggle connecting the brake 
levers. A spring. No. 71. tends to lift the arm 
carrying the pulley. No. 67, and thus hold off 
the brake shoes. When the brake staff is 
turned, it winds up the chain and draws down 
the pulley, No. 67, thereby applying the 
brake shoes to the drum. In this way, t he- 
operator can add hand power to the effect of 

the electric brake and thus produce a greater 
braking action without interfering with the 
automatic operation of the solenoid. 

As appears from Fig. 6, the brake levers, 
No. 55, are double, only the rear member of 
each being shown in Fig. IS. This avoids 
any bending strains on the pivots. The 
levers, No. 60, and No. 68, and the rods, 
No. 62 and No. 70, constituting the two 
toggle systems, are located between the two 
members of each lever, as are also the brake 
shoes, No. 54. The chain, No. 65, extends 
from the pulley, No. 69, to the similar point 
in the brake rigging of the motor at the other 
end of the locomotive, so that the operation of 
either of the brake staffs will apply both 
brakes simultaneously. 

It will be noted that while the hand and 
the solenoid brake mechanism operate entirely 
independent of each other, both apply break- 
ing power through the same levers and wheel. 

Passing now to the features which render 
the locomotive peculiarly adapted for towing 
purposes, it is observed that the drum, No. 
72, Fig. 22, on which the cable, No. 73, Fig. 
5, is wound, is located midway between the 
ends of the locomotive and above the upper 
member, No. 3, Fig. 5, of the side frames, 
so that the cable can be led off on either side 
of the machine and through a wide range of 
angles to the line of travel. The hub, No. 74, 
Fig. 22, of the drum is pivoted to the hub, No. 
75, of the spider, No. 76. which in turn rotates 
upon the upper portion of a massive, tubular, 

Fig. 20. Rear View of a Traction Motor Unit 

vertical cylindrical column. No. 77, rising 
from a pedestal, No. 78, Fig. 25, secured to 
the base plate or floor, No. 79, Fig. 7, which 
is supported upon the lower members, No. 4, 



of the side frames. The upper portion of 
the pedestal is held in a brace, No. 80, Fig. 
24, which is shown as a heavy X-shaped 
casting, fastened to the upper members, No. 
3, of the side frames and to two of the cross 
beams, No. 2. This brace fits the pedestal 
just below the shoulder, No. 81, Fig. 22, on 
which the hub, No. 75, is stepped. 

The spider; No. 76, Fig. 22, supports a 
circular rim, No. 82, which has a horizontal 
upper surface, No. 83, and a flange, No. 84. 
On the surface, No. 83, is secured a flat 
smooth bronze ring, No. 85, and a second 
brass ring, No. 87, similar to the first, lies 
on top of a steel ring and is secured to a 
flanged follower, No. 88. Sixteen studs, No. 
89, project up from the rim, No. 82, through 
holes in a horizontal flange of the follower 
and are encircled by strong springs, No. 90, 
which abut between the flange and nuts, No. 
91, on the studs and press all three rings 
tightly together. The steel ring, No. 86, is 
secured to lugs, No. 92, on a flange, No. 93, 
projecting downward from the winding drum, 
No. 72, so that the rings constitute a friction 
clutch between the spider and the drum. 

Inside the flange, No. 84, on the spider is 
secured a large internal gear, No. 94, with 
which mesh two driving pinions, Nos. 95 
and 96, Fig. 7, secured respectively to two 
upright shafts, Nos. 97 and 98. Step bear- 
ings, Nos. 99 and 100, are provided for these 
shafts in the base of the pedestal, No. 78, 

Fig. 21. Plan View of the Cable Guiding Devices 

Fig. 7, while arms, Nos. 101 and 102, Fig. 

6, projecting from the upper portion of the 
pedestal just below the brace, No. 80, Fig. 

7, constitute guide bearings, Nos. 103 and 

104, for the upper portions of the vertical 
shafts. A worm gear, No. 105, Fig. 7, is 
clutched to the shaft, No. 97, and is driven 
by a worm, No. 106, on the shaft of an elec- 
tric motor, No. 107, bolted to the base, No. 
79, of the locomotive. This gearing is pro- 

Fig. 22. Cross Sectional View of the Cable Guiding 
Devices taken through Line x — x of Fig. 21 

tected by a casing, No. 10S. A bevel gear, No. 
109, is keyed to the upright shaft, No. 98, 
and meshes with a bevel pinion, No. 110, on 
the shaft of an electric motor, No. Ill, 
fastened to the base. 

The motor, No. Ill, with bevel-gear pinion 
is used for driving the drum at a high speed 
when coiling the cable that has been cast 
off, and it remains permanently in gear. The 
other motor, No. 107, with worm-gear drive 
is used for taking in the cable when it is 
under load, and the drum operates as a wind- 
lass or capstan. 

Due to the greater gear reduction, it 
operates the drum at a much slower speed, and 
consequently with motors of approximately 
equal size, a greater force may be exerted on 
the tow-line than would be possible with the 
lower speed reduction which is used with the 
high-speed coiling motor, No. 111. The 
worm-gear drive is disconnected from the 
drum when not in use. To accomplish this 
a clutch is provided, having one member, No. 
112, Fig. 7, splined to the shaft and the other 
member, No. 113, attached to the hub, No. 
114, of the worm gear, which is sleeved on the 
shaft. A lever, No. 115, Fig. 26, fulcru'med 
to a lug, No. 116, on the arm,- No. 101, is 
pivoted to the hub of the clutch member, No. 
112, and its other end is attached to the mov- 
able core of a solenoid, No. 117, which is 
connected in the controller circuit of motor, 
No. Ill, so that whenever the circuit of the 
latter is closed to coil up the cable rapidly, 
the solenoid will lift its core and also lever, 
No. 115, thus throwing out the clutch of the 
winding motor. The first point of the con- 



troller which operates motor, No. Ill, 
raises the clutch and on the second point the 
motor starts. 

The guide which directs the cable, as it is 
paid out or wound up, is mounted so as to 

Fig. 23. Cross Sectional View of the Cable Guiding 
Device taken through Line Y — Y in Fig. 21 

revolve on the axis of the drum. It com- 
prises two angularly adjustable portions, 
Nos. 1 IS and 119, Figs. 21 and 22, the former 
being a circular bell which serves as a cover 
for the winding drum. The hub, No. 120, 
Fig. 22, of the bell is journalled on the upper 
end of the column, No. 77, being stepped on a 
shoulder thereon. At one side the housing 
is cut away to admit the cable to the drum, 
and on each side of this opening is bolted one 
end of a frame comprising box-like ends, No. 
121, Fig. 23, connected by two parallel bars, 
No. 122, Fig. 22, one above and the other 
below the opening. Between the bars and on 
either side of the opening are two upright 
guide rolls, No. 123, Figs. 21 and 22, having 
cylindrical faces, and rotating on journals 
held by the bearings in the bars, No. 122. At 
each end of this frame arms, No. 124, Figs. 
7 and 23, extend downward and support two 
rollers, No. 125, Fig. 7, mounted on hori- 
zontal studs, No. 126, Fig. 23, secured in the 
arms. These rollers are adapted to travel 
between the upper and lower flanges of a 

Fig. 24. Plan View 

of the Pedestal 

and Base 

Fig. 25. Vertical Cross Section of 

the Pedestal and Base taken 

through Line z — 2 in 

Fig. 24 

circular channel-iron, No. 127. Fig. 23, which 
ed on top of the side frames con- 
centric with the column. No. 77, and forms a 
track supporting the outer end of the frame, 

Nos. 121 and 122, thus relieving the column, 
No. 77, of the weight. Stops, No. 128, Fig. 
5, are fastened to the top of the channel- 
iron, No. 127, to limit the angular play of the 
guide member, No. 118. They can readily 
be taken off, and the housing can be turned 
until the rollers, No. 123, are on the opposite 
side of the locomotive, after which the stops 
can be attached on that side to limit the 
movement of the housing. 

The other guide member, No. 119, is a 
radial casting having one end turning freely 
on the hub of member No. 118, Figs. 7 and 
21. A cap, No. 129, Fig. 21, is provided at 
the top of the column, No. 77, which protects 
the joint and prevents the guide members from 
accidentally coming off. The outer end of 
member No. 119 is an upright rectangular 
frame, No. 130, in whose top and bottom is 
journalled on a vertical axis a swivel, No. 131, 
carrying two grooved sheaves, No. 132, these 








Fig. 26. Clutch Operating Mechanism for Slow-Speed Winding 

also being led one above the other on hori- 
zontal axes. The edges of these sheaves are 
in close contact, so that their grooves form 
an opening through which the cable, No. 73, 
passes, approximately in line with the middle 
of the guide rollers, No. 123. The frame, No. 
130, is supported by rollers, No. 133, Fig. 
7, running in track, No. 127, and the guide 
member has an angular movement with ref- 
erence to member No. 118, limited by the 
frame No. 130 striking the ends of the frame 
No. 121. When the cable is pulled either 
forward or backward from the middle posi- 
tion, which it occupies in Fig. 5, the swivel 
permits the grooved rollers, No. 132, Fig. 22, 
to move with it, and the guide member, No. 
119, swings also, so that the rollers, No. 
132, continue to support the rope in a line 
with the middle of the rollers, No. 123, with- 
out being themselves subjected to any side 
strain. All lateral strains are sustained by 
heavy guide rollers, No. 123, the cable moving 



up and down between them as it winds on the 
drum. The latter is in the form of a deeply 
grooved wheel, the groove, No. 134, being 
U-shaped. Figs. 27, 28 and 29 clearly illus- 
trate the construction of the equipment just 

Fig. 29 shows the cable guard. This is a 
steel casting having a thickness of only three- 
eighths of an inch. The diameter is four feet 
six inches and the circular flange is 17 inches 
deep. This casting was pronounced to be 
beyond the possibilities of the ordinary 
open-hearth furnace by a number of steel 
foundries. They were eventually produced, 
however, in the contractor's electric furnace, 
where it was possible to intensify the heat, 
thus making the metal flow more rapidly. 
No failures occurred. With the exception 
noted, all the other principal steel castings 
for these locomotives were produced at the 
plant of the Wheeling Mold & Foundry 
Company, Wheeling, W. Va. 

In order to resist the tendency of the loco- 
motive to tip over when an excessive load 
comes on the cable, a stout rack-rail, No. 
135, Fig. 7, is, as previously mentioned, laid 
between the traction rails of the track, and 
two horizontal flanged wheels, No. 136, are 
arranged between each pair of wheels, No. 1 1 , 
and engage the opposite sides of the rack. 
These wheels are carried on heavy bars, No. 
137, whose inner ends are pivoted at No. 138, 
Fig. 5, to the base of the machine, so that 

One of the most important parts of the 
locomotive is the "slip-friction" device con- 
sisting of two special alloy rings, mounted on 
the spider, as has been previously explained. 

Fig. 27. Complete Assembled Windlass and Base 

Between these rings a steel disk is fastened 
to the rope drum, and the amount of tension 
on the tow-line is adjusted by the pressure 

Fig. 28. Parts of the Windlass and Base and a Partial Assembly 

Fig. 29. Guard for the Towing Cable 

the bars can move horizontally. Their outer 
ends are engaged by strong springs, No. 139, 
which afford the necessary flexibility for 
smooth operation. 

between these three disks, and is obtained 
by tightening the spiral springs on the clamp- 
ing ring. In order, therefore, to make the 
slipping tension of the tow-line proportional 



to the pressure between the friction disks, 
a rubbing surface having an absolutely con- 
stant coefficient of friction is essential. In 
order to find such a metal, certain tests were 
made as indicated by the curves in Fig. 30, 
which is self-explanatory. The low-friction 
metal, having a friction coefficient of 0.1, is 
practically constant under all pressures and 
condition of the surfaces, and therefore was 
selected for the work. This metal also showed 
but very little difference in friction coeffi- 
cient between starting and running. The 
results of the special tests were furthermore 
amply verified by the final test of the friction 
disks of each machine under the full rated 
tow-line pull of 25,000 pounds by means of 
the dynamometer testing outfit shown in 
Fig. 31. All 40 machines were given this slip 
test 25 times from each cab and all passed 
the government requirements not to exceed a 
variation of five per cent above or below the 
normal of 25,000 pounds. 

In connection with the slip test, further 
data on the slow-winding motor was ob- 






■TiOh T~r 


SM« a "- J ^ - 




J0 twW 



^Kep*-"- | 














ZP-' n SjU 


[3 *'"? °'' 



Dn Mel siartinS) tfate 

r : 

" - J' ■ j-. 


; ■ 



. II 4-1 

Mil - 

,,-vton t 

Aetal^Kunniti^) Lubricated witnuil. 


\C " 

30 40 50 60 70 80 90 100 110 1Z0 130 140 150 
Lb. Pressure Per Sq Inert. 

Fig. 30. Curves Showing the Results of Tests to Determine 
the Proper Friction Metal 

tained, as furnished in curves shown in Fig. 
32. The winding motor is a 20-h.p. (one- 
hour rating), three-phase, high- torque, squir- 
rel-cage type, induction motor controlled 

Fig. 31. 

Dynamometer and Stand for Testing the Towing 
Pull of the Locomotives 

from a drum-type reversible controller in 
either of the two cabs. From the curves it is 
seen that the motor has ample power to take 
care of any sudden pull on the tow-line up 
to 40,000 pounds, which is well above the 
normal requirement of 25,000 pounds. The 
speed of winding is at the average rate of 12 
feet per minute. 



C. .1 






c c 

& o 


W u 





2^000 100 fc 

£ 90 5 

x -1 

20,000 40 80 400 

JO 60 300 

?0 40 200 


ZO 100 


















































C 1 



\\ 1 1 1 



H P 

50 60 70 80 90 100 


Fig. 32. Characteristic Curves of the Windlass Motors 



The rapid-coiling motor is permanently 
geared to the drum and is of the same type, 
size, and capacity as the winding motor, 
and is subjected to its maximum load when 
accelerating the heavy drum to the high speed 
required for coiling or paying out the rope, 
this being 16 times the slow- winding speed at 
full load, or about 200 feet per minute. 

The slow-winding and the rapid-coiling 
motors are operated by similar controllers 
and the circuits electrically interlocked so as 
to prohibit application of power to either 
motor unless the controller of the other motor 
is in the "off" position. 

Each of the two main traction motors has 
a rating of 75 h.p., and is of the slip-ring induc- 
tion type, operated by a system of contactors 
with a master controller in each cab. The 
motors, by means of the change in gearing 
from straight traction to rack-rail towing 
previously described, drive the locomotive 
at a speed of two miles per hour when towing 
and five miles per hour when returning idle. 
These motors act as induction generators 
running above synchronous speed when the 
locomotive is passing down the steep inclines 
and thereby exert a retarding brake effect to 
keep the speed uniform. Speed tractive effort 
and efficiency tests were made with results 
as plotted in the curves of Fig. 33. 

The curves in Fig. 34 give some interesting 
data on the time of acceleration of ships in 
the lock chambers. These values have been 
obtained from certain tests and theoretical 

calculations based on data given by several 
well-known authorities. 

For determining the resistance of ships iff 
deep open water, the following formula is 








v ^ 



** ° TJ 



'■ , 

« N "* 




% %9 







■ " 



















A ~~ 






_J '"' n o 







A v - 







200 400 



nput 2 Mo 






Fig. 33. Characteristic Curves of the Traction Motors 

given by Captain Charles W. Dyson in his 
work, "The Estimation of Power for Pro- 
pulsion of Ships:" 

1-85 5 jji y\ 
R=fSV + T 

-L \ -I » ' 

i[iiniini/|i i in ii in ii inn/if 

-- — _: XU 

: 7^ 


± -+- 4 


</ T 

ti^r *i/ ' rj 


n \>y ± _ _,dLiSi ? 

$>* _ WyE- si 

"^ 'i'r~ j 

& _t & 

y*ir ?fl 

S'V -l ti>Y- 

. . -j,e 3 S - - -i-5 E 

\' TT^r 

TTjS^L^ - "S 

<t% -t- " ' " :<X 

(^1-1 ?-° 

.'. >>' 

Wr* T T^ P 

•/ ~SZt 

■+&&', ^ ■>£ 

$ 3& 

+ Ao5g2''T x - S 

' ■»/ . Syr 

jOC^ + 8 s 

-' '< 

m. ' ' J!* 


& ~3&r ^ i 

i- - -a? 

/ ' 



t ' "X- 


j 5 

-4- — -H- S>« 

t"""i"t ~ 8 


1 f- i — ' * 

J! . I j JT "T a 

-E.-C - -H -»?+- -1- - o 

* iS & ^ 

.1 -U - - - 4i 1 ' 

t ' J3~ S -3 

-F -3- T _ 3I S 

r I Tt s — ----*> 

"=f • i T ., I 

Thousand Tons Dispto 


Fig. 34. Curves Showing the Time Required for the Acceleration of Ships 




R = Resistance in pounds. 
/= Surface friction coeffi- 
cient, varying from 
0.008 to 0.009. 
5 = Wetted surface in square 


1 ' = Speed in knots per hour. 

b = Form-factor, varying 

from 0.35 for fine 

long ships to 0.50 for 


D = Displacement of ship in 

L = Length of ship in feet on 

load water line. 
This formula gives, as stat- 
ed, the resistance in deep open 
water, and it is well known 
that this is greatly increased 
when the ship is passing 
through narrow channels, due 
to the reaction of the water 
on the side walls and bottom. 
This additional resistance 
may be found by the follow- 
ing formula, given in a report 
by the State Engineer of New 
York on the proposed Barge 
Canal (see Engineering 
Record, June 29, 1901) : 


r = ratio -. 


canal section 

<]N3 HldON 

midship boat section 

In order to obtain the total 
ship resistance the value of R 
previously obtained should be 
multiplied by the value ob- 
tained for R\. 

Space does not permit of a 
detailed description of the 
locomotive control apparatus, 
but a fairly good idea will be 
had by reference to the dia- 
gram of connections shown 
in Fig. 35. 

Figs. 36 to 43 are of interest 
in showing the progress of the 
work during the construction 

The contract for the loco- 
motives was awarded to the 
General Electric Company at 
Schenectadv. N. Y., U. S. A., 
May 24, 1913. Shipment was 



made of the first machine January 
12, 1914, and the total shipment of 
the forty locomotives was com- 
pleted November 6, or at the very 
high average rate of one locomotive 
per week. The maximum rate 
of production was, however, even 
higher, twelve locomotives being 
completed in nine weeks' time. 

The interest and untiring energy 
of the factory employees engaged 
in this work demand particular 
notice. The men individually 
made it their task to accomplish 
a maximum each day to meet the 
urgent needs of the Panama Canal 
and evinced a striking spirit of 
patriotism and pride in the carry- 
ing out of their share of the big 

The locomotives have a net 
weight of 86,000 lb. and a gross 
shipping weight of 92,500 lb. They were 
mounted on specially designed skids and 
shipped by rail to New York, where they were 
loaded on board ship, as deck cargo, by means 
of aMerritt-Chapman 125-ton floating derrick. 
Fig. 40 shows the loading on the S.S. "Ancon," 
which in this case carried six locomotives to 
the Isthmus. 


The towing locomotives as described and 
illustrated possess the following operating 

(1) When towing, the speed can be accel- 
erated from zero to two miles per hour. 

(2) When running idle, the speed can be 
accelerated from zero to five miles per hour, 
permitting return trips at increased speed. 

(3) The windlass will pay out or wind in 
cable at the low rope speed and at the full 
tow-line pull of 25,000 lb. either with the 
locomotive running or at rest. 

Fig. 36. A Portion of the Assembly Floor of the Contractor's 
Showing a Locomotive Truck Partially Assembled and 
Additional Finished Material 


Fig. 37. 

Front View of a Traction Motor Unit with ! 
Journal Box Disassembled 

Fig. 38. A View of the Locomotive Shown in Fig. 38, but Taken from the Opposite 

Side, with Covers and One Cab Removed Showing the Controllers 

and Front of One Control Panel 






(4) The windlass will pay out or coil 
in cable at the high rope speed with tow-line 
taut either when the locomotive is running or 
at rest. 

(5) The windlass is equipped with a 
safety friction device which is adjustable to 
any predetermined value of tow-line pull. 


The first impression may be gathered that 
these machines are somewhat complicated, 
but considering their many functions and 
great flexibility to perform them, it must be 
agreed that the design is peculiarly simple. 

The locomotives have fully demonstrated 
in actual operation that the requirements 
contemplated by the Engineers of the Isth- 
mian Canal Commission under Circular 650 
have been successfully met. 

During the first three months of com- 
mercial operation of the Canal, from August 
15 to November 15, 1914, the cargo trans- 
ported through the Canal and towed through 
the locks by the locomotives amounted to 
1,079,521 tons. 

During the fiscal year ending June 30, 1914, 

Fig. 42. A View of a Locomotive Crated and Mounted on a 

Flat-Car Ready for Shipment to the Steamship that 

was to Carry it to Canal Zone 

the Panama Railroad carried 643,178 tons 
of through freight between the two seaboards, 
and in the preceding fiscal year 594,040 tons. 
From this it is seen that between six and seven 
times as much cargo is passing over the 
Isthmus now as passed over this route when 
goods were transhipped by rail. 

Fig. 43. A Towing Locomotive on the Test Track in the Contractor's Yard 




Part I. 

By j. P. Minton 

Research Laboratory, Pittsfield Works, General Electric Company 

This is the first of a series of articles dealing with "electrophysics" that we propose to publish during 
the year. These articles will not all be written by the same author. The present contribution on the cathode 
rays and their properties forms an interesting introduction to the subject, and will be followed next month by 
an article on the "electron theory." It is hoped that these contributions will give a useful outline, in simple 
language, of a subject which we feel is of great interest and importance to the engineering fraternity. These 
articles originated in a series of papers presented before the Electrophysics class of the Pittsfield Section 
of the A.I.E.E. They are being revised and amplified by the authors for our columns. — Editor. 



The purpose of this article is threefold; 
first, to demonstrate experimentally that 
there are small negatively electrified particles 
of "something" which are called electrons 
or corpuscles; second, to show that the 
properties of these particles are entirely 
independent of the substances from which 
they come, and, therefore, lead us to the 
fundamental conception of matter; and third. 
to give us a working knowledge of the elec- 
trons in order that we may pursue our future 
study on the electron theory and its applica- 
tions. In the succeeding articles, we shall 
develop the electron theory of electric con- 
duction through solids and gases, and apply 
it to a number of different phenomena. 

We shall consider in the present article : 

1. Historical review (1859 to 1S92). 

2. Experiments on cathode rays, and the 
conclusions derived therefrom which lead us 
to the fact that there are small negatively 
electrified particles called electrons. (a) 
Chemical; (b) Heating, (c) Mechanical; (d) 
Electrical; (e) Magnetic; (/) Experimental 

3. Experimental determination of the 
charge (e), the mass (m), and the velocity 
(v), of electrons. 

4. The constancv of the ratio — , and its 

significance on the fundamental conception 
of matter. 

">. The origin of the mass of the electron, 
and the variation of this mass with the 
velocity of the electron. 

6. Distinction and relation between 
mechanical and electromagnetic mass. 

7. Summary and conclusions. 


In the preparation of this historical review 
Prof. J. J. Thomson's book on "The Con- 

duction of Electricity Through Gases" has 
been freely made use of. 

Cathode rays were discovered by Pluecker 1 
in 1859; he observed on the glass of a highly 
exhausted tube in the neighborhood of the 
cathode a bright phosphorescence of greenish- 
yellow color. He found that these patches of 
phosphorescence changed their position when 
a magnet was brought near them, but that 
their deflection was not of the same nature 
as that of the rest of the discharge. He 
ascribed the phosphorescence to currents of 
electricity which went from the cathode to 
the walls of the tube and then retraced their 
path for some unknown reason. 

The subject was next taken up by Pluecker's 
pupil, Hittorf, 2 to whom we owe the dis- 
covery that a solid body placed between a 
pointed cathode and the walls of the tube 
easts a well-defined shadow, whose shape 
depends only upon that of the body, and not 
upon whether the latter be opaque or trans- 
parent, an insulator or a conductor. 

This observation was confirmed and ex- 
tended by Goldstein, 3 who found that a well 
marked, though not a very sharply defined 
shadow was cast by a small body near the 
cathode, whose area was much greater than 
that of the body. This was a very important 
observation, for it showed that the rays pro- 
ducing the phosphorescence came in a definite 
direction from the cathode. If the cathode 
were replaced by a luminous disk of the same 
size no shadow would be cast by a small 


'Pluecker. Pogg. arm., 107. p. 77. 1859; 116. p. 45, 1862. 

-Hittorf, Pogg. ann., 136. p. 8. 1869. 

'Goldstein Berl. Monat., p. 284. 1876. 

"Varlev. Proc. Row Soc. xix, p. 236. 1871. 

sCrookes, Phil. Trans. Pt. 1. 1879. p. 135. Pt. 2. p. 641, 1879. 

'Hertz Weid., ann.. xlv. p. 2S. 1892. 

T Bancroft. Jour. Franklin Inst.. Feb.. 1913. 

'Millikan. Phys. Review, Vol. 32, p. 349-397, 1911; Aug., 1913. 
pp. 109-143. 

J. J. Thomson. (1) Corpuscular Theory of Matter. (2) Con- 
duction of Electricity through Gases. These two books will be 
found helpful. 



object placed near it, for though the object 
might intercept the rays which came nor- 
mally from the disk, yet enough light would 
be given out sideways by other parts of the 
disk to prevent the shadow being well marked. 
Goldstein, himself, introduced the term 
" Kathodenstrahlen " (cathode rays) for these 
rays, and he regarded them as waves in the 
ether, a view which received much support in 

A very different opinion as to the origin 
of these rays was expressed by Varley, 4 and 
later by Crookes, 5 who advanced many 
weighty arguments in support of the view 
that the cathode rays were electrified particles 
shot out from the cathode at right angles to 
its surface and with great velocity, causing 
phosphorescence and heat by their impact 
with the walls of the tube, and suffering a 
deflection when exposed to the magnetic 
field by virtue of the charge they carried. 
The particles in this theory were supposed 

all are familiar. A diagram of the tube 
is given in Fig. 1, an explanation of which 
follows: Suppose the vacuum in this tube 
has been reduced to 0.006 m.m. of mercury, 
or six microns, and that a static potential 
of, say 15 kv. is applied between the cathode 
(c) and the grounded anode (a) ; the negative 
terminal being connected to (c). A discharge 
will pass through the tube due to the applied 

(a) Now let us see what happens from a 
chemical point of view. First, we shall 
notice a great number of phosphorescent 
patches or spots of light on the glass wall 
over the distance (c) (d), Fig. 1. The color 
of these patches depends on the nature of the 
glass ; thus with soda glass the light is yellow- 
ish-green, with lead glass it is blue. These 
spots can be made to move over the surface 
of the glass by means of an electric or mag- 
netic field without affecting the nature of the 
discharge. This will be made clearer later 

Fig. 1. Cathode Ray Tube 

to be of the dimensions of the ordinary mole- 
cules. The discovery made by Hertz 6 that 
the cathode rays could penetrate thin gold 
leaf or aluminum was difficult to reconcile 
with this view of the rays, although it was 
possible that the metal when exposed to a 
torrent of negatively charged particles, itself 
acted like a cathode and produced phosphor- 
escence on the glass behind. This view, 
however, is not startling since radio-activity 
has been developed, for here we have particles 
going through metals much thicker than gold 

During the past 20 years the workers in 
the field have increased wonderfully, and 
include such men as Weidemann, Schmidt, 
Van't Hoff, Drude, Lorentz, Aberham, Ein- 
stein, J. J. Thomson, 0. W. Richardson, 
Kaufmann, Comstock, Millikan, and many 
other men. No attempt will be made to 
follow their work, but, in a general way the 
results of their experiments which lead us to 
the conception of an electron will be given. 


For these experiments let us consider the 
discharge in a cathode ray tube with which 

in this article. The phosphorescence noted 
is evidently due to something striking the 
glass at these particular places, rather than 
to any wave motion of light, for this could 
not be made to move over the surface in the 
manner described below. 

We also note that there is a violet-reddish 
colored stream of "something" which ap- 
pears to come from the center of the cathode 
(c), and extends over a distance of perhaps 
three inches toward (d), depending on the 
conditions of the experiment. This stream 
is perhaps from one-sixty-fourth inch to 
one-eighth inch in diameter; the larger 
streams being observed the greater the 
pressure in the tube, up to perhaps 12 or 15 
microns. If we put in the tube a diaphragm, 
(d) Fig. 1, of some material, say brass, with 
a hole through it about one-sixty-fourth 
inch in diameter, part of this stream will be 
intercepted while the rest of it will continue 
until it strikes the screen (5) at the point (p). 
This is shown by the fact that, if the screen 
is made of potassium bromide, there will be 
a round bluish-yellow colored spot about one- 
thirty-second inch in diameter on the surface 
of this salt. It appears, therefore, that this 


Salt on the 
Screen (s) 


K Br 
K I 



Cathode Rays 










White— (greenish?) 











stream of "something" starts from the 
cathode (c) and moves to the other end of the 
tube very rapidly, for we can detect no 
difference with the eye in the time of ap- 
pearance of the stream at (c) and at (p). 
The name " Kathodenstrahlen " (cathode 
rays) was given to this stream of "something " 
by Goldstein in 1876. The phosphorescence 
produced by these rays is a chemical phenom- 
enon as is shown by the above table taken 
from the works of W. D. Bancroft. 7 

The following is another table taken from 
the same reference. 

Light Color 

Pb 50 4 -f-cathode ravs - Blue 

Pb + = PbO None 

PbO + S0 3 =Pb SO,..- White 

Pb + (XH,)« S 2 Os=Pb SO, + (\H,), 50 4 Blue 

Zn SOi-f-eathode rays Bluish-white 

Zn+O =ZnO Green 

Zn + S0 3 = Zn SO, Green 

Zn + (XH,) 2 S?O s =Zn SO, + (.NH,h 50 4 Bluish- white 

Cd 504 + cathode rays Yellow 

Cd + = Cd O Yellow 

Cd + S0 3 = Cd SO, Yellow 

Cd + {NH,h 5 2 8 = Cd S0, + (NH,) SO, White 

Considering the first table, suppose that 
solid A T aCl is precipitated out of a solution 
of XaCl. This action consists in Na and CI 
ions uniting to form XaCl which is precipi- 
tated when the solution becomes over satu- 
rated. If this formation of XaCl from its 
ions is observed in a dark room, a bluish-white 
phosphorescence will be observed. Now, this 
is the color caused by the action of cathode 
rays on XaCl. So we conclude that the 
cathode rays cause NaCl to split up into its 
ions, and the immediate combination of these 
ions give off the bluish-white light that we 
observe. Similar remarks apply to the other 
salts listed in the table. With reference to 
the second table, we see that cathode rays 
cause lead and zinc sulphates to break up 
into zinc, lead, and sulphate ions, and the 
reverse action emits the light of the color 
stated. Particles of zinc and lead have been 
found where the rays fell on the sulphates 
of these metals. Sometimes the reverse 

action is very slow as in the case of KBr 
which requires several hours to reach the 
initial conditions again. In the case of other 
salts, like calcium tungstate, the reverse 
action is practically instantaneous. The 
former is called phosphorescence and the 
latter is called fluorescence. 

If the screen (s) is an oxidized copper 
plate, the cathode rays soon cause a bright 
copper colored spot to appear: that is, these 
rays exert a reduced action. The rays also 
affect photographic plates as was shown in a 
recent article (Comptes Rendus, 158, pp. 
1339-1341, May 11, 1914) by A. Dufour on 
"The Cathode Ray Oscillograph." He ob- 
tained photographs corresponding to a deflec- 
tion of the ray stream of 1 mm. in 3 X 10~ 6 sec. 
It was necessary to use very strong rays to 
obtain such results. 

(b) Having considered some of the chem- 
ical effects produced by these rays, let us 
next take up their thermal effects. These 
have been investigated by J. J. Thomson, 
E. Weidemann, Ebert, Ewers, and others, 
all of whom have found that these cathode 
rays heat bodies on which they fall. If the 
rays are concentrated by using a spherical 
shell cathode, platinum may be raised to 
incandescence, thin pieces of glass fused, 
and the surface of diamond charred. A 
simple example will give some idea of the 
amount of energy carried by these rays. 
It must be stated first, however, that these 
rays are composed of negatively charged 
electrons as will be shown later. So let n 
be the number of electrons striking the sur- 
face in unit time, m the mass of the electron, 
and v its velocity, the energy E given up to 
the body in unit time by the electrons is 

E=-r n m v- where this is the total kinetic 

energy transformed into heat energy on 
striking the surface. If e is the electronic 
charge, then the current carried by these 

ravs is I = ne, or n = 


Hence E- 

1 , m 

: 2 7 7* 




Now 10~ 5 amperes is a fair value for I, and if 

d = 5X10 9 cm. per sec., — = 6X10 _S , then 


£ = yX10- 5 X6Xl(T 8 X25X10 18 = 7.5X10 7 

ergs. Since one calorie equals 4.2 X10 7 ergs, 
E equals approximately 1.7 calories. All of 
this energy does not produce heat, but some 
is used in producing Rontgen rays, second- 
ary cathode rays, and some electrons are 

(c) Mechanical effects of cathode rays 
are also important. A typical example of 
this was carried out by Crookes in 1879. 
He placed the axle of a very light mill with a 
series of vanes on glass rails in a vacuum tube. 
When the discharge passed through the tube 
the cathode rays struck against the upper 
vanes and the mill rotated, traveling toward 
the positive end of the tube. If the potential 
was reversed, the direction of rotation also 
reversed, showing that the cathode rays were 
now moving in the opposite direction. Since 
the upper limit of the momentum given to 
the vanes by the rays is of the order of magni- 
tude of 10~ 2 dynes, this alone cannot account 
for the rotation of the vanes. It was shown 
later to be largely due to the heating effect 
produced by the cathode rays on the side on 
which they impinged. 

Another exceedingly important mechanical 
effect is that these rays pass through a thin 
gold leaf, and where the velocity is quite 
high they have passed through 1 mm. of 
aluminium. This is equivalent to passing 
through 250 miles of molecules if they were 
two inches in diameter. This had an impor- 
tant bearing on the final acceptance of the 
view that cathode rays consisted of very small 
particles and were not a wave motion of any 

It may be mentioned here, as stated under 
/, that the fact that the cathode rays came 
from the negative terminal in a definite 
direction was a further proof that these 
rays consisted of particles of "something." 
The name electrons was given to these par- 
ticles in 1890. It was also shown that these 
electrons came directly from the cathode, 
otherwise they would not have been inter- 
cepted by an object placed in their path. 

(d) The electrical effects produced by 
these rays- show conclusively that they are 
particles of matter. First of all, if a beam of 
light passes through the air and falls on the 
wall, the spot of light will not be affected 
by presence of a magnetic or an electric field 

near this wall. Furthermore light does not 
possess an electric charge, for electricity 
always associates itself with matter. 

It has been shown that cathode rays move 
from the negative to the positive terminal, 
and must, therefore, possess a negative 
charge. This is further affirmed by the fact 
that, if a direct current potential is applied 
to the set of quadrants QQ, Fig. 1, the phos- 
phorescent spot on the screen (s) will move 
in such a direction as demanded of a negative 
charge by the fundamental laws of electricity. 
Bodies upon which the rays strike acquire 
a negative charge. Those experimental facts 
will suffice to show that the cathode rays 
possess a negative charge, and must be 
associated with small particles of matter. 
The charge on these particles is a certain 
definite quantity (as will be shown later), 
and one never finds an electric charge which 
is not a multiple of this fundamental and 
elementary unit of electricity. This, then, 
indicates that the cathode rays are atomic in 
structure and the electricity resides on these 
small particles. 

(e) Cathode rays are deflected by a mag- 
net when the field is not parallel to the direc- 
tion of motion of the electrons. This also 
indicates that we are dealing with concrete 
particles which carry a negative charge as 
shown by the direction of the deflection. . 

(/) To summarize: It has been shown 
that there are such things as cathode rays as 
revealed by the effects they produce, viz., 
chemical, thermal, mechanical, electrical, and 
magnetic. Furthermore it has been shown 
that they are atomic in structure, being com- 
posed of small negatively electrified par- 
ticles. These are the accepted conclusions 
of the scientific world. 

III. DETERMINATION OF (e), (m), and (v) 

Let us, now, investigate the properties of 
these electrons somewhat in detail. We shall 
first determine experimentally their charge 
0), their mass im), and their velocity (v). 
Referring to Fig. 1, suppose that we have a 
stream of electrons passing through the tube 
and that they produce a phosphorescent spot 
at (p). Now suppose to the quadrants QQ 
is applied a steady known potential which 
causes the spot to move to a new position 
(p 1 ) . Let (h) be so large that h = h l for our 

In moving over the length h, the electrons 
fall through a distance D due to the electric 
field applied to QQ. As in the case of falling 



bodies, we have: D = -at 2 , where a is the 

acceleration of the electrons due to the 
electric force E applied to the quadrants, and 
t is the time required to move over the path 

h Now t = — , and the force on an electron 


equal Ee = Ma or a = — . If we substitute 

these values for t and a in the equation for 

, . „ 1 Eeh 2 e 2 D v 2 
D, we obtain D = - „ , or — = 

2 mv 2 



In this equation you will find the three 
fundamental quantities of an electron, viz., 
e, m, and v. If we know v, we can, therefore, 

obtain — which is called the specific charge. 

To do this a magnetic field is superimposed 
upon the electric field at the quadrants QQ 
in such a way as to balance the effect pro- 
duced by the latter field. Now Rowland 
showed experimentally that the force on an 
electron due to the magnetic field is H e v sin 8, 
where 6 is the angle between the electric and 
magnetic fields and H is the strength of the 
latter. Making = 90 deg. (or sin 0=li we 

have, therefore. H c v = Ee or b=tj. Both E 

and H can be easily measured, so that » is 
known. This velocity never exceeds 3X10 10 
cm. per sec. the velocity of light: 10 9 is a fair 
velocity for the electrons. In radio-activity v 
is perhaps 2.5 X 10 10 cm. per sec. and in the 
cathode ray tubes it is about 5X10 8 cm. per 
sec. Putting this value of v in the equation 

for — , we obtain — = 1S00 X 10 4 . Now — for 
m m m 

the hydrogen ion is 10 4 , so that the specific 
charge of an electron is about 1S00 times as 
large as that of the hydrogen ion, the smallest 
particle of matter yet known. To settle 
this point we need only to measure the charge 
(c) of the electron. At least eight different 
methods have been used for this purpose. 
These methods are two radio-active, one 
Brownian movement, one radiation, two 
"cloud formation," one Zemarm effect, and 
the famous oil-drop method of R. A. Milli- 
kan 7 of the University of Chicago. It is 
well to note that the values given by these 
methods agree within three per cent of that 
given by Millikan, who is absolutely certain 
of his value to 0.1 per cent. The other in- 

vestigators do not claim any such accuracy 
for their results. His method of determining 
e is briefly as follows : 

He immersed a brass vessel in an oil bath 
to maintain a constant temperature. Within 
this vessel were two parallel metallic plates 
1.6 cm. apart. The air pressure in the brass 
chamber could be varied at will by means of 
pumps. X-rays could be passed through a 
glass window so as to ionize the air between 
the plates, thus causing free electrons to 
exist in this space. Now, by means of an 
atomizer extremely small (0.0005 cm.) drops 
of oil could be sprayed between the plates, 
and when either an ion or an electron stuck 
to the drop it would acquire a corresponding 
charge. He applied a potential to the plates 
and so could move the oil drop up and down 
at will. He could also detect, by a change in 
the velocity of the drop, when a new charge 
attached itself to the drop ; which was viewed 
by means of an optical system. He used the 
following formula to calculate the charge on 
the drop due to (n) elementary charges: 

47T (9rf\ 1 ( 1 V/ fa+tfrA 
1 3\,2/ \&Qr-P))\ F ) 


where (77) is the coefficient of viscosity of 
air, (a) the density of the oil, (p) the 
air density, (vi) the speed of descent of the 
drop under gravity, and (i' 2 ) its speed of 
ascent under the influence of an electric field 
of strength F. All of these quantities were 
known to 0.1 per cent. The equation was 
based on three assumptions, viz., the drag 
which the medium exerts upon a given drop 
is unaffected by its charge ; neither distortion 
due to the electric field nor internal con- 
vection within the drop modified appreciably 
the law of motion of the drop; the density of 
the oil drops is independent of their radii 
down to 0.0005 cm. Millikan not only 
showed that these assumptions were justi- 
fiable, but their effects were not present 
at all. 

By means of the above equation he ob- 
tained a series of values for (e„), and taking 
the lowest one he found all the others to be 
exact multiples of it. This value we naturally 
accept as the elementary charge. He has 
carried out a great number of tests under 
various conditions as regards size, tempera- 
ture, pressure, and gives e = 4.774 ± 0.009 X 
10~ 10 electrostatic units. 

Now, this value is the same as that carried 
by a hydrogen ion, which must, therefore, 
carry one of these elementary charges. We 
are led then to the experimental conclusion 



that the mass of the electron is 


of that 

of the hydrogen atom, which until now was 
the smallest particle of matter we had known. 
We are forced, therefore, to the fact that 
matter is still further divisible than we were 
led to believe by the atomic hypothesis. 
The mass of the electron is, therefore, m = 
4.8X10- l0 X3X10- 10 . . 

1SX10 7 ' whlch glves '" = 8 >< 

10 -27 grams. This is true for velocities which 
are not very close to 3X10 10 
will be shown shortly. 

cm. per sec. as 


The ratio 

i0 (A 


(e), and (w), have been 

measured for many different kinds of gases, 
solids, and elements under various conditions. 
In every case all of these quantities were 
constant (except for velocities near that of 
light) and entirely independent of the sub- 
stance from which they were obtained. 
Consequently, the electron is a fundamental 
unit of electricity and matter, and all matter 
must have it as one of its constituents. The 
other constituent or constituents, as the case 
may be, must be matter which acquires a 
positive charge by losing electrons and gain 
a negative charge by addition of electrons. 
If there are positive electrons, however, this 
conclusion need not be true; but the scientific 
world has tried in vain to discover them for 
the past 15 years. Until they are discovered, 
we must content ourselves to build up a 
theory of matter with the electrons as a 
basis. This is the so-called electron theory 
of matter, and, since they always possess a 
negative charge, they form the basis of the 
electron theory of electricity. Both of these 
theories, which will be developed in the next 
article, are "subject to change without 

Fig. 2 

notice." As to the nature of positive elec- 
tricity we know nothing except it must exist. 


It is a"well-known fact that when a current 
of electricity flows through a wire a magnetic 
field is set up in the space around it. Simi- 

larly, a charged body moving through space 
sets up a magnetic field in the surrounding 
space. Hence, if we have a charged body, it 
will require more energy to set it in motion 
with a velocity (i>), than would be required 

*; — x 

Fig. 3 

to set it in motion with the same velocity 
if it were not charged. In the first case the 

energy required is - M v* plus the magnetic 

energy, and in the second case it is - M v 2 . 

Since the velocities are equal, it follows that 
the charged body apparently has a greater 
mass than the uncharged one. This fictitious 
mass due to a magnetic field surrounding a 
moving, charged body is called the electro- 
magnetic mass of the body. 

Now let us determine the electromagnetic 
mass of a charged sphere of radius a, moving 
through space with a velocity (v), where (n) 
is not too nearly equal to the velocity of 
light (c). Let 0) be the charge on the sphere. 
Rowland has proved experimentally that the 
magnetic force at a point p. Fig. 2, due to the 
charge (e) moving along OX is: 

tt _e v sin d 

H ~ ~1*~~- (1) 

The energv density at (p) due to the mag- 

H 2 
netic field alone is Dm = ; 

or, by equation (1) : 

8 (pi = ir) 

c- v- sin 2 
8T(pi=w) r k 




Changing Fig. 2 to Fig. 3, the volume 

abcdefgh = -~ r sin 6 r d r d 6. 

Multiplying this equation by 4, we have 

4 a b c d e j g h = 2 (w) r 2 sin 6 d d dr (3) 
This equation gives the volume of an element 
of an imaginary sphere in the space surround- 
ing the charged sphere. The magnetic 
energy in this volume is, therefore, by equa- 
tions"^) and (3). 

(A) E H = ^ff x20r)r'sin 6 dd dr 


(A) E H = 


e 2 v"sm 3 6 dd dr 
4r 2 

Hence, the total energy due to the magnetic 
field in the space surrounding the sphere is 

E H 


'± a Jo 


c-v-si>! 3 6 


1 <iird 

dd dr 


dd dr 


) sin 3 6 dd 


1 3 cosO (sin 2 6 + 2) 


e' i- 


The total energy, Et. therefore, possessed 
by the moving charged sphere is 

. e 2 ;•'-' 

E T =1 2 .1/ v 2 +l 3 


E T =l/2 

[ (M+ Hl 

where M is the mechanical and 


2 e- 

electromagnetic mass. If we assume an 
electron to be spherical and that it acts as 
though the charge were located at the center, 
equation (5) applies to it just as it does to 
the sphere. 

In 1901 Kaufmann determined experi- 
mentally the value of the quantity in brack- 
ets in equation (5) for different velocities 
of the electrons. The results he obtained are 
illustrated in Fig. 4, where (c) is the velocity 
of light. From this curve we see that the 
apparent mass of an electron becomes infinite 
when v = c, and that it changes very little up 
to 2X10 10 cm, sec. If the curve were con- 

tinued to v = o, it would cut the ordinate 
v = o at M, which is the mechanical mass of 
an electron. Since this latter mass does 
not change with v, it follows that the electro- 
magnetic mass must increase very rapidly 
when v=c (means v approaches c as a limit). 

This is also the conclusion at which one 
arrives from a purely mathematical con- 
sideration. The deduction is quite compli- 
cated and I shall not frighten you by giving 
it in this article. The equations show, how- 
ever, that when vi=c there is a weakening in 
both the electric and magnetic fields in the 
regions of a o b and cod, and an increase in 
the regions a o d and b o c, see Fig. 5. When 
v = c, the fields are zero except in the plane 
g o h, where they are infinite. Hence when 
v = c, the electromagnetic mass of an electron 
becomes infinite. Since by far the greater 
part of the mass of an electron is electro- 
magnetic, it must necessarily possess very 
little inertia, even up to velocities comparable 
with those of light. 

( ,_^ 

i j _n j m 

& j 1 — X-ttt-tif 11 

£ 4==: _ix_iii_i± xxtrxLT 
iniiiiiiii" t ii in id 

I I Li)! It J Lilt 

IIIIIII IJJT j i_ ! Tj~ i ~ 

— r-H — i — i — H — ^K 

— i — i — I — , — i — i — i — ! — i — i — -\ — 

— ----- -j — h±--—3B: 



1,/D"> Z*/0'° 


Fig. 4 


Equation (5) must be modified to cor- 
respond with this change in mass according 
to the following equation 

E B =\ -lUl+j (v) 2 

3 a 




where / {v) = l when v = o, and / (v)= in- 
finity when v = c. In this connection the 
following table, which was taken from J. J. 
Thomson's Corpuscular Theory of Matter, 
p. 33, will be interesting. 

From equations (5) : 









Fig. 5 

Velocity of 

Ratio of 

Total Mass 

to that of 

a Slow 



2.36 X10 10 cm /sec. 
2.48 X10 10 cm /sec. 
2.59 X10 10 cm /sec. 
2.72 X10 10 cm /sec. 
2.85 X10 10 cm/sec. 



A consideration of these values obtained 
by Kaufmann will show that the mass of an 
electron is almost wholly due to the magnetic 
energy in the space surrounding it. Thomson 
concludes that the mass of an electron is 
entirely electromagnetic in origin. This 
conclusion is not justifiable for velocities 
below 2.59X10 10 cm/sec. as is seen from the 
table above so that the electron must possess 
some mechanical mass even though it may 
be an extremely small per cent of the 

Assuming the mass (m) of an electron to be 
wholly electromagnetic, we can calculate its 
radius (a) as follows: 










:= — 





We have seen that — =1.SX10 7 , and that 

e=10~ 20 electromagnetic units. Substituting 
these values in (7), we obtain a = lCT 13 cm. 
approximately; the radius of an atom or 
molecule is about 10~ 8 cm. 

The relation and distinction between me- 
chanical and electromagnetic mass in the 
above discussion have been pointed out. In 
addition, it has been shown that: 

(1) Electromagnetic mass must have weight ; 

(2) Electromagnetic mass = constant X me- 
chanical mass. Some theoretical physicists 
even go so far as to assume all mass is electro- 
magnetic. On account of insufficient time, 
however, the line of argument leading up to 
these conclusions is not given. 


From the information here given we must 
conclude that there are small negatively 
electrified particles called electrons, the prop- 
erties of which are entirely independent of 
their source. We have seen that these elec- 
trons exist in matter as well as separated 
from it like cathode rays, Beta particles from 
radioactive substances, in gases where a 
discharge of electricity is passing, etc. 

In concluding this article the author 
wishes to say that he has endeavored to 
briefly present the experiments upon which 
the electron theory is based, and has not 
developed it at all, simply suggesting it. 
He has also endeavored to familiarize you 
with the electronic conception sufficiently to 
develop the theory and apply it to various 
phenomena in the succeeding articles. 



By J. F. Layng 
Railway and Traction Engineering Department, General Electric Company 

The author deals with some of the important considerations governing the selection of car equipments 
for city and suburban service. By making an analysis of the pressures of wheel treads, he determines the 
method of mounting the motors that will give the maximum adhesion available for traction. It is then 
possible to determine the equipment that will be most suitable to operate on severe grade conditions, and 
also to obtain the best schedule speed on all rail conditions. — Editor. 

The purpose of this article is to designate, 
in a general way, the facts concerned in the 
selection of car equipments to meet the 
varied conditions which confront railway- 
engineers when purchases are to be made. 
The situation can best be covered by an 

In general, there are at least six different 
classes of service for equipment at the present 
time, viz.. city, interurban, elevated, subway, 
steam railway terminal electrification and 
main line railway electrification. Insofar as 
the present discussion is concerned, only 
city and interurban service will be considered. 

The electrical equipment of city car service 
may be divided into two classes, viz., two and 
four-motor equipments; of trucks, in general, 
there are three types, viz., single trucks, 
maximum-traction trucks, and double trucks. 
The number of combinations that can be 
made when applying power to the car with 
these elements is surprising. The proportion 
of the total car weight on the driving wheels 
largely determines the schedule possibilities 
and the grade-climbing capacity of a car. 
With the single-truck, two-motor equipment, 
all the weight is on the driving wheels so 
that this combination would be ideal were it 
not for the fact that the demands of seating 
capacity and riding quality put limitations 
on the single-truck car that usually make it 
necessary to have double-truck equipments. 
With the double-truck car there are many 
complications which arise when selecting a 
distribution of power for the driving axles. 
A selection which will give uniform weights 
on the wheel treads will, of course, give the 
ideal car, for it will reduce wheel slippage 
to a minimum under all conditions. Dur- 
ing the period of acceleration there is a 
shifting of the car-body weight so that 
there is a lesser weight on the front center- 
plate than on the rear center-plate. The car- 
body weight assumed for all the double- 
truck cars considered in this discussion is 
20,000 pounds or 10 tons, and this mass is 
assumed to be accelerated at the rate of 1^2 

miles per hour per second. There is-a retard- 
ing pressure of 137 pounds per ton, or a total 
of 1370 pounds, due to acceleration. It is 
assumed that the center of this mass is 24 
inches above the center-plate, and that the 
king-pins are 20 feet between' centers. When 
the car is accelerating there is a shifting of car 
body weight around the center of the mass, 
which, with the car body as described when 
accelerating at IJ2 miles per hour per second, 
gives 9S63 pounds weight on the front 
center-plate and 10,137 pounds on the rear 
center-plate. The same shifting of weights, 
but in different proportions, takes place on 
the trucks. This action is independent of 
the position of the motors on the trucks. 

Later on it will be shown that with a four- 
motor equipment and motors "inside hung" 
it is possible to secure the nearest approach 
to equalization of the weights on the wheel 

h / »l 

Fig. 1 
L = Distance between truck centers. 
H = Distance of center of gravity of car body above center 

W —Weight of car body in pounds. 

T = Retarding pressure of car body during acceleration. 
P = Pounds pressure transferred from front truck to back 

truck at king-pin. 
Id —Weight on front center-plate. 
Wi = Weight on rear center- plate. 
pi = Pounds pressure transferred from front to rear axle 

(center-plate load). 
pi = Pressure transferred from front to rear axle (truck load). 
P = pi+pi- 

W = Weight on front axle exclusive of truck weights. 
W = Weight on rear axle exclusive of truck weights. 
b = Weight of truck. 
/ = Retarding pressure of truck during acceleration. 

The question is frequently asked — "What 
equipment shall we buy, two-motor or four- 
motor?" To this question a direct answer 
cannot be given; it is a question of judgment. 
The answer is determined by the amount of 
wheel slippage that is allowable On this 



account, generally speaking, double-truck 
two-motor equipments are not satisfactory 
where the grades exceed five per cent, or 
on a bad rail such as is produced by snow, 
sleet, mud, or leaves on the track. Wheel 
slippage is the factor which usually decides 
whether or not four-motor equipments are 

The combinations of weight distribution 
and the weight on the individual wheel treads 
during the period of acceleration present a 
very interesting problem. The maximum 
schedule will of course be maintained by the 
equipment which has the most nearly equal 
weight distribution on the wheel treads. 

In addition to the many combinations of 
motor mounting for a single car, we have trailer 
operation to consider and also the effect of 
these trailers on wheel slippage, both on level 
track and also on grades. An analysis of the 
weight distribution on single-motor cars in- 
dicates clearly the reason for usually selecting 
four-motor equipments when trailer opera- 
tion is to be considered. This analysis of 
weight distribution shows why, when the 
grades to be negotiated are more than five 
per cent it is the general practice to use four- 
motor equipments for double-truck cars. 

There are at least twelve to fifteen dif- 
ferent combinations of mounting motors on 
the different types of trucks in general use. 
For 'the present purpose, twelve combina- 
tions will be considered. With each of these 
there is a different weight on wheel treads 
during the acceleration period. In arriving 
at the values given in Fig. 2, it is assumed: 

1st. The car is accelerating at one and 
one-half miles per hour per second on tangent 
level track. 

2nd. The single-truck car weighs complete 
10 tons, has a ten-foot wheel base, and the 
center of gravity of the car is 24 inches above 
the center-line of the axles. 

3rd. All double-truck car bodies complete 
with car-body equipment weigh 20,000 pounds 
each with no live load. 

4th. Double-truck cars have trucks weigh- 
ing 6000 pounds each. Those having inside- 
hung motors have 72-in. wheel-base and those 
having outside-hung motors have 54-in. 
wheel-base. The distance from the surface 
of the center bearing-plate to the center-line 
of the axle is 12 inches, the wheels are of 33- 
in. diameter, the center of gravity of the truck 
with motors is 16^ i n - above rail for motors 
shown in Fig. 2, and the distance between the 
truck centers is 20 feet. The weights and 
lengths of each end of the car are equal. 

5th. There is a slight downward tilting 
of the car body and trucks during the accel- 
eration which will make a slight variation 
in the figures as given, but as these variations 
are inappreciable and would complicate the 
explanation, no allowance has been made for 
this variation. The rotative effects of dead 

Direct /on of cor motion — 
Tot ol cor body weight ZOOOOIb 

]^ 3gn ,A I 











S283 4S76 



+ 228 




/O 66 4 lb. 94741b 
Weights os given are for iveightson wheels of each axle. 
Fig. A. Specific example of how to determine the distribution 
of car weights when the center of gravity of the trucks 
and motors is taken as 18 in. above head of rail 

Distribution of Car Weights on Wheel Treads 

During Period of Acceleration (Four 

Motors Outside Hung) 

Car body weight. 20.000 lb. 

Truck weight, 6000 lb. +two motors, each 2000 lb. = 

10.000 lb. total. 
Wheels 33 in. diameter. 
Wheel-base 54 in. 
Truck centers 20 ft. or 240 in. 
Center-plate 12 in. above center-line of axle and 28 } ■• in. 

above wheel tread. 
Center of gravity of car 24 in. above center-plate. 
Center of gravity of trucks with motors IS in. above 

wheel tread. 
Car acceleration 1 K miles per hour per second. 
Ninety-one lb. required to accelerate one ton one mile per 

hour per second XI H =137 lb. 
Car body weight transferred from front to rear center- 
137X20,000 lb. X24in. ,_, ,, ... 

P ' ate= ^00 1b.X240 1b. =137 lb ' Wh ' Ch B ' VeS 

9S63 lb. weight on front center-plate and 10.137 lb. on 

rear center-plate. 
The 9S63 lb. on front truck has transferred weight from 

front to rear axle as follows: 

2000 X 54 

'=356 lb. 

transferred weight. 
The 10.137 lb. on rear truck has transferred weight from 

front to rear axle as follows: 

2000 X 54 


transferred weight. 
Within the truck and motors there is also weight trans- 

f erred from front to rear axle as follows: 

2000 X54 

= 22Slb. 

axles have not been considered, as the varia- 
tion would be slight. 

Based on these facts, the outlines in Fig. 
2 have been made. 

The formulas from which the weight pres- 
sures were calculated are derived in the fol- 
lowing, in which the diagram in Fig. 1 and 
its key are used as a base. 




The retarding pressure of the car body will 

T = 

137 IT 

<ap- W U&o oW 

1 392 3 SSOB €s&2 tg 77' 

&@y t®« 


6 30O Z2&29 

1377/ SOS2 '3 929 S2eB '377/ Sa32 

'2339 6739 ';;« 6602 >0&03 92SS 69tX »»J"B 



90+3 --eoss I07++ 3"& 


&2ST *07+6 O'-a 


- 1 t=3Tsa 

1 fe-B — a L 

■0SO4 9360 1049? 964/ 

'03 GO 9S04- 

0if?eCT/O* Of fit- L CA/?3 *- 

r/ft/#£-S G'V£# fif£ V*£lGH7SOff WHEEL Ttf£ffPS WHEN 
accELEtRT/NG /^ M F? H P£F S£C0MI> 

Fig. 2. Twelve Methods of Mounting Motors on Trucks, 

Showing Relative Weight Distribution 

During Acceleration 

The transfer of weight from the front 
center-plate to the rear center-plate is found 
by taking moments^ about a point X as 

137JI7/ L 

2000 2 

137 WH 
2000 L 

The transfer of car-body weight from the 
front axle to the rear axle is determined, as 
follows : 

-^K=pih+pik=pi(.k+h)= : pil. 


The transfer of truck weight from the front 
to the rear axle is determined thus : 

, 137 b 137 b h 

t = ~^X7^ —^T^TT = Pit 

137 bh 


yi 2000 I 
ti * TK , 137 bh 
Then ^ = 2T+2000l" 

And the distribution of wheel pressure will 
be as follows: 

Front truck 
W=W 1 l j+p 

Rear truck 
W' = W^-p 

W=W s l f+p 

Considering the weights on the individual 
wheel treads, it will be seen that with the 
single-truck car, shown in Fig. 2, there is a 
noticeable shifting of the weight from the 
front axle to the rear axle during acceleration. 
The wheel-tread weights for the maximum- 
traction trucks are also somewhat different 
from those we would naturally expect from 
the static weights (in all the assumptions 
made for maximum-traction truck calcula- 
tions, it has been assumed that the static 
weights on the driving wheels are 70 per cent 
of the total). It is a natural conclusion that 
with four-motor equipments the nearest to 
equal distribution of weight on wheel-treads 
under all conditions would be procured. 
However, by comparing diagrams 1 1 and 12, 
Fig. 2, it will be noted that the inside-hung 
motor arrangement shows a considerable 
improvement over four-motor equipments 
with outside-hung motors. All of the weights 
as given in the different diagrams are for cars 
accelerating at 1 J/£ miles per hour per second 
on straight level track. When accelerating 
on grades the values will be somewhat dif- 
ferent, but not as much as one would naturally 
be led to expect. 

Another feature of these calculations which 
should be considered in connection with the 
weights is that cars which have long plat- 
forms on one end only have not been con- 
sidered. There are so many variations in 
this respect in cars operated in regular service 
that it would be impossible to make general 
statements covering this condition. Dimen- 
sions and weights on both ends of the cars 
have therefore been assumed to be equal. 

After the question of deciding how many 
motors are to be used on a car, the next 
factors to consider are how to get the greatest 



amount of work out of the motors per pound 
of weight, how to secure the motor that will 
use the least power, and at the same time 
to obtain an equipment at a price that will 
be justified by the results of these savings. 

It is a universally accepted fact that a 
ventilated motor will have a greater work 
capacity per pound of weight than a non- 
ventilated motor. The past few years' ex- 
perience has led all of the truly progressive 
engineers to specify that the motors which 
they are about to purchase must be ventilated. 
It is also a generally recognized fact that by 
using field control the work capacity of a 
motor, that is properly designed, is increased. 
When considering the use of field control, the 
service should be carefully reviewed to see if 
the increased cost and the complications of 
the field control are warranted by the savings 
in power and weight of the extra control 
apparatus required. The cost of a motor 
designed for field control is but slightly 
greater than that of one for full-field running 
only, but there is also an increased cost in the 
control and car wiring for the former type of 
motor. There are practically twice as many 
field connections as are found in the ordinary 
full-field equipments. These extra con- 
nections make the locating of trouble more 
difficult. All of these factors must be taken 
into consideration. Practically all the sav- 
ings secured by field control are made during 
the period of acceleration, and, since this is 
the case, rapid acceleration decreases the 
power saving. Another way of partially 
expressing this idea is that where steps are 
infrequent the saving is proportionately small. 
On account of the complications in wiring and 
control but comparatively few field-control 
four-motor equipments have been installed. 
However, for two-motor equipments many 
purchases of field-control equipments for 
frequent stop city service have been made. 

During the past three years there has been 
an increased amount of interest exhibited 
regarding 24-inch wheel equipments for city 
service. Motors that are particularly effi- 
cient and well constructed have been designed 
for use with these equipments. Due to the 
decreased weight of the wheels and trucks 
as well as to the reduction in the weight of 
the motors, this subject has engaged active 
study. In addition to the weight savings there 
has also been an innovation in control which 
consists of a change in the standard motor 
circuit connections so that three running 
speeds are obtained. With this combination 
of control there is a considerable saving in 

power consumption. In some cases a saving 
of seven or eight per cent may be expected. 
It is necessary, however, in connection with 
this control to carefully analyze the service, 
for experience has shown that the heating 
is not equally divided among the four motors 
of the equipment. Referring again to the 
reduction in the weight of trucks for these 
equipments, it can be stated that for the 
standard 33-inch wheel equipment (which 
it was formerly the practice to use) these 
trucks would weigh approximately 12,000 
pounds per pair, and that it is now possible 
to purchase trucks with 24-inch wheels which 
will weigh but 8000 pounds per pair. The 
weight of an individual motor which was 
formerly designed to operate with the 33-inch 
wheels would be approximately 2000 pounds 
while the 24-inch wheel motor complete 
weighs but 1750 pounds. It can therefore 
be seen that a saving of 4000 pounds can be 
made in the trucks alone and in the motors 
1000 pounds additional, making a total re- 
duction in car weight of 5000 pounds. This 
weight saving is something that cannot be 
ignored. There have been a number of 24- 
inch wheel equipments purchased during the 
past year and the results of their performance 
will be followed very carefully. In all proba- 
bility, within the next year or two, some very 
pertinent facts will be available. 

In the selection of interurban equipments 
the same considerations in regard to motor 
distribution apply as have been mentioned 
for city cars. Practically all equipments in 
this class of service employ four motors, and 
it is the usual practice for the motors to be 
inside-hung. There has not been a general 
adoption of field control for interurban work 
due to the fact that as a rule the stops, as 
compared with city service, are relatively 
infrequent and therefore the savings which 
can be made with city equipments is not so 
apparent in the interurban equipments. 

When purchasing any large number of new 
equipments the savings which may be pro- 
cured with higher voltages than 600, which 
has been the standard in past years, are very 
attractive due to the savings which would 
seem to be possible. However, an analysis 
of many of the existing interurban road con- 
ditions develops the fact that these interurban 
roads also supply city service to a large 
number of small towns through which they 
pass, and that on these city equipments very 
extensive changes would have to be made. 
That is, the electrical equipment would all 
practically have to be renewed on these cars. 



Provided this change is not made the line 
must be so sectionalized that the town sec- 
tions would still operate on 600 volts, and 
after all of the savings and the additional 
costs have been taken into consideration, it 
has been found that in a great many cases 
the change was not warranted. Of course, 
if these properties were entirely new the prop- 
osicion would be a much different one, and in 
all probability the installation of the higher 
voltages would be more than warranted. 

A large number of interurban roads have 
practically reached the limit of their earning 
power as they are securing all of the busi- 
ness which there is at the present time, 
and the only additional business which can 
come to them is through the natural increase 
of business due to the growth of the com- 
munity. This has lead the management of 
these properties to carefully consider and to 
estimate the cost of entering into the business 
of hauling car-load freight. As a result it 
would not be surprising if a large number of 
roads purchase locomotives in the near 
future in order to increase their earnings per 
mile of track. To select the motors for this 
service, it has usually been the practice to 
start with locomotives weighing approxi- 
mately 40 tons. Sometimes 
these units are of the regular 
locomotive type and some- 
times of the baggage-car type. 
These units as a rule are 
equipped with four 125-h.p. 
motors. On some properties 3 
this business has grown to g'^ 7 ^'" 
such an extent that it is 
necessary to use 60-ton loco- 
motives which are usually 
equipped with four 225-h.p. 

In the selection of an equip- 
ment for either city or inter- 
urban work, it is necessary to 
have a very definite picture 
of the work to be done. Very 
careful consideration of the 
actual work that is desired to 
be accomplished should be 
given by the management of 
the railway company; and 
when working up data for 
a proposition, it is very dangerous for the 
railway company to put leeway in the figures 
which they give to the manufacturing com- 
pany. When certain equipment is recom- 
mended by a manufacturer the railway com- 
pany can always place dependence upon it, 

zoo 27 
180 18 
160 16 

140 ^u 


60% 6 
40 4 

zo z 

for in the recommendation the manufacturer 
has of necessity already embodied a certain 
margin of safety. 

If there has been an allowance made 
earlier by the railway company as to schedule 
speeds, stops per mile, duration of stops, 
etc., and then the manufacturing company, 
ignorant of the previous allowances, also 
makes additional ones in each of these factors ; 
the equipment will be larger than is actually 
required to perform the work, and this would 
be caused simply by doubling the allowances. 
The number of stops and the duration of 
stops which are made per mile by any equip- 
ment are very deceptive. The only way in 
which this info-T>- +: ~n can be obtained is by 
actual observation and careful records. The 
problem is not a difficult one, but is just 
an actual statement of the facts. If 
these few statements were given careful 
consideration frequently, considerable of the 
extra figuring and time of all parties con- 
cerned would be saved. The securing of 
accurate service data insures the purchase 
of the proper size equipment. This means 
an equipment which will give satisfac- 
tion and also will be secured at a reason- 
able cost. 

^""" ~"~~-^ J 

: v *>'" " *S 

X^ VU- 

Y xV 

-Zk ^S 

7 V Sv 

- v S * ± 

-■ -7 ^ its 

= ■ 7 *-, S\ I 

7 "I t\ i 

/ S ' 

/ \ ^ i 

.' ~~ ">>P P 

(-— i_ v ^ 

6 8 10 12 IA 16 18 20 22 24 26 23 30 32 34 36 38 40 
Tin-i& - Seconds 
Fig. 3. Speed-time and Energy Curves for 20-ton Cars 

Comparison of energy consumption for 

slightly different schedules. 
Service assumptions: 

Two-motor equipment — 50-h.p. motors. 
20 tons total car weight. 
o50 volts average 
7-75 stops per mile. 
1H m.p.h.r.s. accelerating and braking 

20 lb. per ton friction. 

Dotted line curves indicate: 

Maximum schedule 9.6 m.p.h. No 

Energy consumption 145 watthour per 

Full line curves indicate slightly de- 
creased schedule, 9.45 m.p.h., coasting 
about 30 per cent distance. Energy con- 
sumption 122 watthour per ton-mile. 

Showing a saving of 19 per cent in energy. 

In a general way, it can be said that at- 
tempting to run a faster schedule than normal 
is very expensive. An illustration of what 
happens to power consumption under this 
condition with 20-ton cars is shown by the 
speed-time and energy curve shown in Fig. 


3. This is based on 7.75 stops per mile. 
With 9.6 miles per hour schedule speed, 
145 watthours per ton mile will be used. Bv 
slowing this schedule to 9.45 miles per hour, 
122 watthours per ton mile are required; a 
saving of 19 per cent. It would seem that 
in the selecting of the equipment and in the 

laying out of schedules a considerable power 
saving could be made by being a little more 
reasonable in the running time. Of course 
there is a natural tendency to run the highest 
possible schedule speeds at the present time, 
which has been brought about by the recent 
increase in "platform" wages. 


By B. W. Jones 
Industrial Control Department, General Electric Company 

Since at the present time p.c^Mcally all motors are shipped from the factory complete with the proper 
starting rheostat, the information given in this article will be of use mainly to the factory designing engineer; 
yet at the same time formula; of the kind will be of great assistance in some cases of temporary motor instal- 
lations where it is necessary to regulate the starting characteristics of the motor. The author deduces simple 
formula; for the resistance steps for maximum and minimum given values of accelerating current with given 
number of rheostat divisions and value of internal motor resistance, for shunt, series and induction motors 
singly, and for two and four series motors in series-parallel. The theory on which the method is based is then 
given, and a concrete example for each of the cases assumed is worked out. — Editor. 

In commercial practice it is necessary to 
calculate a large number of starting rheostats 
for the shunt, induction, and series types of 
motors, and therefore it is essential that a 
short method be used. Of the methods 
available, the one described in the present 
article has been found in practice to give 
remarkable satisfaction. 

First, either the maximum or the minimum 
accelerating current peak is assumed, together 
with the number of rheostat divisions and the 
internal motor resistance. If it is a series- 
motor resistance, then the speed character- 
istic curve of the motor is necessary. During 
acceleration the successive divisions of re- 
sistance are short-circuited as fast as the 
current decreases to a fixed assumed value, 
and all the current peaks are of equal value. 

It should be noted that if all values are in 
percentages it will cause less confusion. 
Therefore, all of the series-motor formuke 
will be expressed on that assumption. 

Ri = Total 

current = 

resistance to give minimum 


Total resistance to give maximum 


= Internal resistance of motor. 

; r*; r-i\ ?C etc., = Resistance of the 

successive divisions of the rheostat. 
= Number of rheostat divisions. 

maximum acceleration current 

current : 

X = Ratio 

Ai = 

Z : 



100 V-hr 

100 V 

acceleration current 


S, I, 


V =Line volts. 


i i = Minimum acceleration current = -=-. 


Si = (For series motors) speed correspond- 
ing to II. 

Is = Maximum acceleration current = 

For Shunt or Induction Motors 

I. Assume that the minimum accelerating 
current, the number of rheostat divisions, and 
the internal resistance of the motor are known. 

Log X 





5j = ( For series motors) speed corre- 
sponding to h. 
5.3 = Speed corresponding to 1.5 h* 


r t = Xri. 

r 3 = Xr2. 
r„ = Xr n -\. 

*See footnote, page 132. 



II. Assume that the maximum accelerating 
current, the number of rheostat divisions, and 
the internal resistance of the motor are known. 

Log A =- logy. 

n = (A-l)r. 

r3 = Xr 1 . 
r„ = Xr „-i. 
It is apparent that if any three values are 
known, the fourth can be found. 

For Series Motors 

III. Assume that the minimum accelerating 
current, the number of rheostat divisions, and 
the internal resistance of the motor are known. 

The following is an empirical formula: 

1 . /?! 

' n + 1 

Then, from the article "Determination of 
Resistance Steps for the Acceleration of 
Series Motors" by E. R. Carichoff and H. 
Pender in the General Electric Review. 
July, 1910, we take 

n = A x (Si- S 2 ). 

r 2 = Zr x . 

r 3 = Zr 2 . 

r n = Zr „_i. 

IV. Assume the same conditions as in III. 

but that there are two motors to start in series- 

For the series position: 

1 . R 



n -f 





Log X 


r 2 = Zn. 
r 3 =Zr 2 . 
Yn =Zr n —i. 

For the multiple position 

Log A = - — log 


n+ 1 

r 2 = Zr x . 
r 3 = Zr 2 . 
r„ = Zr„-i. 

■S t ). 

* Note that the denominator, 1.51 

( 5i - 5» \ 

(Si) '—(Si) 

It is sometimes convenient and slightly more accurate to make 
/a as near equal to li as can be approximated instead of h = 

V. Assume the same conditions as in III, but 
that there are four motors to start in series- 
parallel. Four motors are to be in series at 
starting and two are to be in series for permanent 

With this condition if r = internal resist- 
ance of two motors then the formulas in 
V are to be applied. 

I. From geometrical progressions: 
Sum = r+rX 1 +rX 2 +rX 3 +. . .rX" + r+X"+'- 
(Where n equals two less than the number 
of terms which correspond to the number 
of rheostat divisions.) 

Therefore the last term, R\, which is the 
sum of all the previous terms is: 

X"+ 1 = 


LogA = 





II. If n represents three less than the 
number of terms, then the next to the last 
term is 

R, = rX", 


t v l i R * 
Log A =- log-. 

III. For series motors a modification of 
the above is necessary. Since there are 
already too many variables to solve the 
equation, it is necessary to resort to an 
empirical formula which is accurate enough 
for all practical purposes. 

The motor field flux varies as a function 
of the current, and the current fluctuation is 
a function of the motor's internal resistance 
and the number of rheostat divisions. There- 
fore in place of 

r there is placed 1. 

in + 2) 


LogA = 





IV. The only difference between starting 
one motor from rest to full speed or two 
motors in series from rest to one-half speed is 
that an extra resistance, (r), is inserted in the 
circuit. Since the two motors in series 
attain only one-half speed, the sum of the 
two counter e.m.f. increments are the same 
as those of one motor running full speed when 



the current peaks in each case are the same. 
Therefore, the formula becomes 

Log X = 






\n + 2) 


When the two motors are changed to the 
parallel connection and are accelerated from 
one-half to full speed, then the extra resist- 
ance (r), that was in the series connection, is 
eliminated and the average counter e.m.f. 
increment corresponding to each current peak 
is doubled. Therefore, the formula becomes 

Log X = 


n + l 





V. For the series-parallel operation of four 
motors, the motors are assumed to be con- 
nected as follows: During series operation 
all the motors are in series, and during parallel 
operation two sets each consisting of two 
motors connected in series are across the line. 
Then, if the internal resistance of two motors 
is considered as r, the same formula? as given 
in V hold true. 


Assume: A 25 h.p., 230-volt shunt motor. 
Minimum accelerating current =100 amp. 
Internal resistance of motor = 0.2 ohms. 
Number of rheostat divisions = 4. 

Log X = 






= T log 0.2 5 
= 0.212 
Therefore A'= 1.63. 

Then n=(X-l)r 

to = Xr i 
n = Xr» 

Ti = A> 3 

log 11.5 = 

\- (1.061). 
5 v 

= 0.126 ohms 
= 0.205 ohms 
= 0.334 ohms 
= 0.544 ohms 

Total external resistance 1.209 ohms 
Internal motor resistance 0.200 ohms 

Total resistance 
Therefore as a check 

1.409 ohms 
230 volts 

1.409 ohms 

163 amp. 


Assume : Same conditions as in Example I, 
except that the maximum current is known 

(163 amps.) and the minimum current is to 

be found. 


Log X 


1 , i?2 

=— log — 

n r 

i log w = T log7 - 05= r (0 - S49) 

= 0.212. 

T -3T 

-t- i 




- i 




~T - % x 


s _ 


V - - Z 

X i ,<!+'-. 

H \ =' J' 

J ' 't-^c'f''^.. 1 ;<!■- X ! 

\ i "** t-jsp"' 

\ ' 4-i^ 7 *■? 

\ ! 4- tf£ ^'J- 

i_L .tiz "~0&&. 

•s ~-t— -S*" ^ V 1 

/ ! sf ~~ 7 — ~~ * '° •** 

? ■ £?• ^ — — 

v 7 Jti' ^ 

*^ ,E' 

Z ' ^' • 

. 1 

a IS 

■ SOO 120 

//OO //a 
/ooo ^too soo 

900 %9(. 

800 0. 30 400 

0, 600 * 60 \ 300 

300 \ JO 
S ' 

OO ^ 4C 


300 Jj 30 
200 $ 20 
/OO /O 

/CO 200 300 400 SOO COO 700 aOO 900 /OOO //OO /SOO '300 1400 
TorqtJB- lb at /ft ra^'tjs 

Fig. 1. Characteristic Curves of a 50-H.P., 230-Volt 
Series Motor 

Therefore X= 1.63 (same as in Example I). 
The remaining is the same as in Example I. 


Assume: A 50-h.p., 230-volt series motor. 
Minimum accelerating current = 185 amp. 
Internal resistance of motor = 0.087 ohms. 
Number of rheostat divisions = 4. 

Referring to characteristic curves of this 
motor, Fig. 1, all values should be reduced to 
percentages. The speed and current for 
minimum accelerating values, together with 
line voltage, will always be considered 100 
per cent. 

V =230 volts = 100 per cent volts. 
I\ = 185 amp. = 100 per cent amp. 
Si = 578 r.p.m. = 100 per cent speed cor- 
responding to 100 per cent amp. 
S3 = 487 r.p.m.=S3 per cent speed cor- 
responding to 150 per cent current. 
y = (S, -S 3 = (100-83) = 17 per cent. 




L°g* = ^ lo = 



» + l 


= 1_ 100 

~5 g 1.5X5/6X17 

= 0.135. 
Therefore, A" =1.364. 

. 100X100-100X6 
A\ = 

= 9- log 4.7 = 

A 2 = 

87 X 136.4 


" 11850 


r x = .Ai(Si-S») =0.94(100-87= 12.20 

per cent ohms. 
H=Zr\= 14.80 per cent ohms. 
r 3 = Zr 2 = 17.95 per cent ohms. 
r 4 = Zr3 = 21.75 per cent ohms. 

Total external resistance 66.70 per cent 

Internal motor resistance 6.00 per cent 

Total resistance in circuit 72.70 per cent 

V V 230 1 Ol V, 

R=y-= r^ = 1-24 ohms. 
1\ loo 

Therefore, to reduce the percentages to 
actual values 

12.20 per cent of 1.24 = 0.151 ohms 
14.S0 per cent of 1.24 = 0.184 ohms 
17.95 per cent of 1.24 = 0.222 ohms 
21.75 per cent of 1.24 = 0.270 ohms 


The other variations for two or four motors, 
if worked in per cent volts, current, speed, 
and resistance, are so similar to those just 
given that it is not considered necessary to 
work out an example. 



By Dr. Wheeler P. Davey 
Research Laboratory, General Electric Company 

Two other articles describing the applications of the Coolidge X-Ray tube have appeared in the 
Review, one "Some Interesting Applications of the Coolidge X-Ray Tube" in the August, 1914. number, 
p. 792, the other "An X-Ray Inspection of a Steel Casting" in the January, 1915, number, p. 25. The 
present article deals with the examination of the interior structure of copper castings and presents an inter- 
esting stereoscopic radiograph of a block of porous copper from which, by the aid of a stereoscope, the pores 
can be viewed in perspective. — Editor 

Dr. Weintraub in the February, 1913, 
number of the Journal of Industrial and 
Engineering Chemistry describing boron and 
its compounds says: 

"Boron suboxide, a by-product obtained 
in the manufacture of boron, can be used for 
obtaining high conductivity cast copper. 
Copper cast without additions is full of pores 
and blowholes, and therefore mechanically 
unfit and of very low electric conductivity; 
the removal of the gases from copper by the 
known deoxidizers is liable to give an allov 
containing a small amount of deoxidizer, an 
amount sufficient, however, to lower the con- 
ductivity of the copper very considerably. 
Boron suboxide, however, has the property 
of deoxidizing copper without combining 
with it, as boron suboxide has no affinity for 
copper. Tons of copper are cast now by this 

process, improving the quality of the product 
and at the same time cheapening it." 

In the refining of copper for electrical 
purposes, the electrically deposited metal is 
melted in a reverberatory furnace. A world 
of delicate chemical control is connected with 
this furnace refining. When ready to pour, 
the metal is cast into open iron moulds 
which give a copper pig or bar of about 75 
lb. in weight. 

If the metal were merely melted and then 
poured the casting would be full of blow- 
holes and would be of low electrical conduc- 
tivity. The molten copper is allowed to 
oxidize in the furnace and the oxidation is 
augmented by air blown into the metal. 
When the melt contains five or six per cent 
of oxide, the major part of the other impuri- 
ties have been burned away and the work of 


Fig. 1. Radiograph of a Block of " Unboronized " (Pure) Copper side by side with a Block of " Boronized " 
(Pure) Copper. Note difference of internal structure 

reduction is started. As ordinarily done, this 
consists in the so-called "poling." Green 
sticks are submerged in the molten copper 
and the gases and carbon reduce the oxide, 
and such harmful products as sulphur 
dioxide are driven out of the metal. The 
proper time for pouring is not that represent- 
ing complete reduction of all oxide, as 
it has been determined by experience that 
over-poling also gives a porous inferior 

It was once believed that the copper 
absorbed carbon which in over-poled copper 
caused the rising in the mold and the porous 

condition when cast. Hampe corrected this 
idea and attributed the porous state of over- 
poled copper to the effect of absorbed hydro- 
gen and carbon monoxide. In any case the 
fact remains that if we merely melt copper 
and cast it we get a porous casting, and if we 
thoroughly remove dissolved oxygen by car- 
bon or similar reducing agents, we also get a 
porous casting. 

The use of the boron flux of Weintraub has 
done away entirely with the difficulty of 
obtaining sound castings of high electrical 
conductivity. It seemed interesting to illus- 
trate the effect on the porosity by an investi- 




1 1 


Fig. 2. 

Ordinary Photograph of the Block of 
" Unboronized " Copper 

Fig. 3. 

Ordinary Photograph of the Block of 
" Boronized " Copper 



gation using X-rays. For this purpose some 
high grade copper was melted in the usual 
way and poured into a sand mold to give 
a block 10 by 10 by % inches. Another portion 
was treated with one per cent of the boron 
flux at the time of pouring and was cast in a 
similar mold. These two castings were then 
placed side by side on an S by 10-inch Seed 
X-ray plate, 22 inches from the focal spot of a 
Coolidge X-ray tube and exposed for two 
minutes. The current through the tube was 
2.8 milli-amperes and the potential difference 
across the tube corresponded to a 10-inch 
parallel spark gap between points. The 
resulting radiograph is shown in Fig. 1. The 
copper cast in the ordinary way is seen to be 
full of pores. The cast with the boron flux 
is so perfect that no holes are visible. The 
two castings were then taken to the machine 
shop and a portion of the surface of each was 
machined as smooth as possible. Ordinary 
photographs were then taken, see Figs. 2 
and 3. As was to have been expected from 
the radiograph, Fig. 1, the holes were clearly 
visible in the common copper. In the "boron- 

ized" copper the holes are either entirely 
absent or are microscopic. 

The advantage of the radiograph in experi- 
mental work is obvious. Without the use of 
X-rays it is necessary to machine off layer 
after layer of the sample in order to expose 
to view any hidden defects. Even when this 
is done it remains for the experimenter to 
build up a mental picture of the defects in his 
casting on the basis of what he has seen on 
each of the exposed layers. From the radio- 
graph it is possible to see all of these defects 
at once without destroying the casting. If 
it seems desirable, it is easily possible to make 
stereoscopic radiographs whereby the de- 
fects may be seen in their entirety and their 
depths easily estimated. Such a stereoscopic 
radiograph of a portion of the pure copper 
casting is shown in Fig. 4. This figure should 
be viewed through an ordinary stereoscope. 

In view of the results shown above, the 
X-ray examination of metals as a means of 
metallurgical research seems to have certain 
attractive and desirable features not found in 
other methods and to open a wide field for 
further work. 

Fig. 4. Stereoscopic Radiograph of a Portion of the Block of Unboronized Copper (Actual Size). 
When viewed through a hand-stereoscope this shows the size and relative depths of the pores 



By F. W. Peek, Jr. 

Consulting Engineering Department, General Electric Company 

It has been a long-established fact that geographical altitude (or air pressure) and temperature has a 
very material effect on the value of an insulator's spark-over voltage. Although these influences have been 
recognized, we believe that the following article records the results of the first comprehensive tests to determine 
their amounts. When given sufficient data regarding an insulator, the tabulations and charts in this article 
enable the spark-over voltage to be determined at any temperature and atmospheric pressure. The article 
was presented as a paper at a meeting of the A.I.E.E. in December, 1914. — Editor. 

The following investigation was made to 
determine the effect of air density, and there- 
fore of altitude or barometric pressure, and 
temperature, upon the spark-over voltages 
of leads, insulators, etc. 

The dielectric strength of air decreases 
with decreasing pressure and increasing tem- 
perature; that is, with the relative density 
or with the average spacing of the molecules. 
If the relative density is taken as unity at a 
standard pressure of 76 cm. and a temperature 
of 25 deg. C, the relative density at any other 
pressure and temperature is 
* 3.926 
5 = 273 + * Where 
b = barometric pressure in cm . , and 
t = temperature in degrees C. 
For the uniform field between parallel 
planes the spark-over voltage decreases di- 

rectly with 8. If e is the spark-over voltage 
for a given spacing at 5 = 1, the spark-over 
voltage e x at 5 = 0.5 is 

^1 = 0.5 e. 

The effect is the same for the same value of 
5 whether 5 is changed by temperature or by 
pressure. This has been shown elsewhere.* 
For non-uniform fields, as those around wires, 
spheres, insulators, etc., the spark-over volt- 
age decreases at a lesser rate than the air 
density. The theoretical reasons for this 
have been given, as well as the laws for 
regular symmetrical electrodes, for cylinders, 
and spheres, f 

It is, however, not possible to give an exact 
law covering all types of leads, insulators, 
etc., as every part of the surface has its effect. 

* Law of Corona II. A.I.E.E. Trans.. 1912, p. 1051. 
t Law of Corona II, A.I.E.E. Trans., 1912. p. 1051. and 
Law of Corona III, A.I.E.E. Trans., 1913. p. 1767. 

Fig. 1. 

Cask for Study of Variation of Spark-Over and Corona Voltage with 
Air Density or Altitude 












Deg. C. 


















































LEADS See Fig. 17) 


Fig. 2 Fig. 3 Fig. 4 Fig. 5 












n sy 



















The following curves and tables give the 
actual test results on leads, insulators, and 
bushings of the standard types. The cor- 
rection factor for any other lead or insulator 
of the same type may be estimated with 
sufficient accuracy. When there is doubt, b 
may be taken as the maximum correction. 
It will generally be advisable to take 5 because 

the local corona point on leads and insulators 
will vary directly with 5. This is so because 
the corona must always start on an insulator 
in a field which is locally more or less uniform. 

The tests were made by placing the leads 
or insulators in a large wooden cask 2.1 
meters high by 1.8 meters inside diameter, 
exhausting the air to approximately 5 = 0.5, 
gradually admitting air and taking the 
spark-over voltage at various densities as 
the air pressure increased. The temperature 
was always read, and varied between 16 and 
25 deg. C. The cask is shown in Fig. 1. 

At the start a number of tests were made 
to see if a spark-over in the cask had any 
effect upon the following spark-overs by 
ionization or otherwise. It was found that 
a number of spark-overs could be made in the 
cask with no appreciable effect. During the 
test, the air was always dried and the surfaces 
of the insulators were kept clean.* 

Table I is a typical data sheet. Tables II 
and VI give even values of 5 and the corre- 
sponding measured correction factors. If the 
spark -over voltage is known at sea level or 
5 = 1 (76 cm. bar., temperature 25 deg. C.) 
the spark-over at any other value of 8 may 
be found by multiplying by the corresponding 

* In these tests, corrections have been made for wave shape, 
etc., and the voltages checked by sphere gap. Voltages measured 
by needle gap are incorrect and indicate higher voltages than 
really exist. 

1 £~4.:^-- 










Fig. 2. Arc-Over Voltages at 
Various Densities 











Fig. 3. Arc-Over Voltaees at 
Various Air Densities 


- . - 

■ iV : 


















7 8cr 







Ht -~ T ■£ DENSiT* 

Fig. 4. Arc-Over Voltages at 
Various Densities 


Fig. 5. Arc-Over Voltages at 

Various Densities 



correction factor. It will be noted that in 
most cases the correction factors are very 
nearly equal to 5. 

Fig. 15 is a curve giving different altitudes 
and corresponding 5 at 25 deg. C. If the 
spark-over voltage is known at sea level at 
25 deg. C, the spark-over voltage at any other 
altitude may be estimated by multiplying 
by the corresponding <5, or more closely if 
the design is the same as any in the tables, 
by the correction factor corresponding to 8. 
If the local corona starting point is known at 
sea level, it may be found for any altitude by 
multiplying by the corresponding d. The 
barometric pressure corresponding to dif- 
ferent altitudes is given in Fig. 16. Figs. 17 












Fig. 6. Arc-Over Voltages at 
Various Air Densities 

3 5 cr 



u*- 1 




Fig. 7. Arc-Over Voltages at 
Various Densities 


-■36cm - 








Fig. 8. Arc-Over Voltages at 
Various Densities 

to 20 show the insulators used in these tests. 

As an example of the methods of making 

corrections: Assume a suspension insulator 

string of four units with a spark-over voltage 




Fig. 6 Fig. 7 Fig. 8 

Post Pin 


1.00 1.00 
0.93 0.91 
0.84 0.81 
0.76 0.73 
0.68 0.62 
0.60 0.52 


SUSPENSION INSULATOR, FIG. 9 (See Figs. 18 and 19) 



Number of Units 


2 3 


5 - 



1.00 1.00 
0.93 ii 'in 
0.84 0.80 
0.76 0.70 
0.66 0.60 
0.55 0.50 


Fig. 9. Arc-Over Voltages at 
Various Air Densities 



of 205 kv. (at sea level, 25 deg. C. tempera- 
ture). 5 = 1. What is the spark-over voltage 
at 9000 ft. elevation and 25 deg. C? 

From Fig. 15, the 5 corresponding to 9000 
ft. is 

5 = 0.71. 

SUSPENSION INSULATOR, FIG. 10 (See Figs. 18 and 19) 



Number of Units 




4 ' 







































SUSPENSION INSULATOR, FIG. 11 (See Figs. 18 and 19) 


Number of Units 


















Then the approximate spark-over voltage at 
9000 ft., 25 deg. C. is 

e x = 0.71X205 = 145 kv. 

If this happens to be the insulator of Fig. 
10, the correction factor corresponding to 
5 = 0.71 is found in Table V, by interpolation, 
to be 0.73. The actual spark-over voltage 
for this special case is 

<? 2 = 0.73 X 205 = 1 50 kv. 
The first estimate is on the safe side and close 
enough for all practical purposes. Thus, for 
practical work the correction may generally 
be made directly by use of Fig. 15. 

The spark-over voltage of an insulator is 
100 kv. at 70 cm. barometer and 20 deg. C. 
What is the approximate spark-over voltage 
at 50 cm. barometer and 10 deg. C? 

5l -273X20-°- 94 

= 3.92X50 
2 273 + 10 = 

= 0.61 

ei=100X^ = 65kv. 

If the local corona starting point is known 
at, say, sea level, it may be found very closely 
for any other altitude by multiplying by the 
correction 5. 

The spark-over voltage of insulators will 
vary somewhat from day to day, due to 
humidity. There is also some variation for 
different units. The humidity voltage varia- 
tion on the insulator is possibly as high as 7 


Fig. 10. Arc-Over Voltages at 
Various Air Densities 






4 units 




/ - 








l ' 





(7 cm 






/ ' 



/ , 








Fig. 11. Arc-Over Voltages at 
Various Air Densities 



si | 

10 ! 

5 i 

.5 Q 

■S 6 

.S Q a 

CO z£ 




J units . 





1 1 

6 2*crn^ 



-ItsL-Jl j 





■ 26 












- 1 u 

nil — 








Fig. 12. Arc-Over Voltages at 
Various Air Densities 

per cent, from day to day. Comparative 
tests of different types, when desired, should 
be made at the same time. The humidity 
correction, on the insulator itself, is too com- 
plicated to make and of no practical value. 








o 7000 



3 » 





\ 1 


Si S" ■ 

i J> 


Fig. 13. Arc-Over Voltages at 

Various Air Densities 

Horn gap spark-over. Gap spacing 14 cm. 

Diameter of horns 1.27 cm. 


"^ITrrr, - 




Fig. 14. Arc-Over Voltages at 
Various Air Densities 

Care must be taken, however, to use a measur- 
ing gap unaffected by humidity; that is, a 
sphere gap. 














■ 000 


Fig. 15 


Fig. 16 



By L. C. Porter 

Edison Lamp Works, Harrison, N. J. 

Much has been published regarding the superiority of the drawn-wire tungsten-filament lamp over the 
old carbon-filament lamp in installations on land. This article analyses the conditions which determine 
illumination on shipboard; and shows that the new lamp when installed there displays the same excellent 
qualities that it does on land. — Editor. 

Lighting practice ashore has advanced so 
far and so rapidly during the past few years 
that the public, now, not only appreciates 
but demands good lighting everywhere. As 
a result, ship owners have been made to feel 
this influence; and, in consequence of this, 
a large amount of attention and study has 
been given to the lighting of passenger and 
naval ships. The vast superiority of the 
tungsten-filament lamp over the carbon- 
filament lamp having become universally 
recognized on land, it is only natural that the 
newer lamp should also replace the older 
when used on shipboard. As a matter of 
fact, this is what is taking place. 

The use of tungsten lamps aboard ships 
offers several mechanical advantages in addi- 
tion to those obtained by their use ashore. 
Space and weight are of considerable moment 
on board our modern high-speed passenger 
boats. Reliability is also of great importance. 
Repairs when necessary must frequently be 
made at short notice and without most of the 

and lighter generating apparatus, lighter wir- 
ing throughout the ship, less coal storage space 
necessary, etc. 

It frequently happens, when changing over 
the lighting equipment of a ship from carbon 

Fig. 2. 

Dining Room on the "Adirondack" of the 
Hudson Navigation Company 

Fig. 1. The Saloon on the Steamship "Priscilla" 
of the Fall River Line 

facilities obtainable ashore. The use of 
tungsten-filament lamps, consuming approx- 
imately 1.10 watts per c-p. as against 3.10 to 
3.50 by carbon-filament lamps, means smaller 

lamps to tungsten lamps, that one generator 
may be shut down and held in reserve. 
Smaller and lighter storage batteries for emer- 
gency use are required for tungsten lamps 
than for carbon lamps. 

On certain steamers, people often ride 
solely for pleasure or for pleasure and business 
combined. To assist in drawing their patron- 
age a lighting system must be devised which 
is not only necessary but ornamental as well. 
For this reason all-frosted lamps are almost 
invariably found on passenger vessels. These 
lamps frequently are made further decorative 
by the use of round bulbs and occasionally by 
placing them in elaborate chandeliers or 
equipping them with fancy reflectors. The 
general interior finish aboard ships is white. 
The bright all-frosted tungsten-filament lamps 
harmonize well with white wood work. Artis- 
tic effects are given the most consideration on 
ocean liners and the least on ferry boats. Be- 
tween these two comes the great class of 
coastwise, river, and lake steamers. 



The lighting problems of passenger vessels 
may be divided into five main parts: (1) social 
halls, (2) dining rooms, (3) smoking rooms, 
(4) staterooms, and (5) passageways. In 

Fig. 3. Smoking Room on one of the Old Dominion 
Line Steamers 

addition to these there are certain parts of 
the ship where a relatively small number of 
lamps are used, such as engine rooms, bar 
rooms, barber shops, boiler rooms, freight 
holds, etc. 

Investigation shows that there are two 
general types of steamers in the coastwise, 
river, and lake class, those making compara- 
tively short runs and those making trips of 
several days. On the former, the social halls 
consist of a large well, or opening running up 
one or two decks above the main deck. This 
class of social hall is well illustrated by the 
photograph taken on the Fall River Line 
steamer "Priscilla," Fig. 1. On ships making 
long trips, the social halls are smaller and 
generally but one deck high. Also they are 
usually not lighted quite as elaborately. 
Ceiling lamps are found frequently supple- 
mented by lamps in wall brackets. 

Illumination measurements, taken on a 
large number of ships lighted with carbon 
lamps, showed an average foot-candle in- 
tensity, on a plane three feet above the floor, 
of 1.1 on long trip ships, and 1.5 for the short 
run class. 

An interesting demonstration of the great 
improvement obtained by substituting tung- 
sten-filament lamps for carbon lamps was 
made in the social hall of a boat of the 
short run type. Thirty-eight sixty-watt 
all-frosted carbon lamps, giving an average 
intensity of 1.4 foot-candles for an energy 
consumption of 2.3 watts per square foot, were 

replaced by thirty-eight twenty-five-watt 
all-frosted tungsten lamps. These lamps gave 
2.7 foot-candles illumination with an energy 
expenditure of 1.0 watt per sq. ft. 

Dining rooms are very similar on each class 
of ship. Round-ball enclosing globes located 
on the ceilings were considerably used. On 
one ship the 16-c-p. clear carbon lamps in 
enclosing globes were replaced lamp for lamp 
by 25-watt clear tungsten lamps. The carbon 
lamps gave 1.1 foot-candles for 1.5 watts per 
sq. ft., while the tungsten lamps in the same 
fixtures gave 1.4 foot-candles for 0.6 watts 
per sq. ft. Fig. 2 shows a section of the dining 
room of the Hudson River Line steamer 

Smoking rooms have a fairly high ceiling 
on the short run ships and a low one on the 
long trip class. Fig. 3 shows the arrangement 
employed in the smoking-room of a steam- 
ship on the line of the Old Dominion S. S. Co., 
which illustrates the latter type of lighting. 
All-frosted lamps located overhead are usually 
used. The average illumination 3 feet above 
the floor was found to be 1.25 foot-candles for 
an energy consumption of 2.09 watts per 
square foot. 

In passageways a low illumination is all 
that is needed. This is frequently obtained 
by small lamps located on the ceiling spaced 
about 10 feet apart. The average intensity is 
about 0.S foot-candles. 

Fig. 4. 

Stateroom on the "City of Montgomery," 
Savannah Line 

The staterooms are generally rather poorly 
lighted, having but one lamp located in the 
center of the ceiling. Great improvement is 
obtained by locating this lamp over the center 
and one foot out from the mirror, thus allow- 



ing comfortable shaving, etc., and at the same 
time giving good general illumination in the 
room. A stateroom on the Savannah Line 
steamer "City of Montgomery," having a 

lamp for lamp by 40-watt medium efficiency 
tungsten lamps. The former gave an average 
of 1.13 foot-candles, on a plane 3 feet above 
the floor, with an energy consumption of 1.96 

60 Watt Carbon Lamps 

40 Watt Mazda Lamps 

Figs. 5 and 6. New York City Ferryboat "Richmond" 

lamp at the center of the ceiling and a portable 
lamp near the berths, is shown in Fig. 4. 

The lighting of ferries was found to be very 
uniform. In practically all cases the cabins 
were lighted by carbon lamps, in one- or two- 
light fixtures located on the walls over the 

tk ** "" 

- ,-:; 

*-» -diH 

t - - 

' : «-> 1 







Palm Garden on the "Victoria Louise", of the 
Hamburg-American Line 

seats. In a few cases round-ball all-frosted 
lamps were in use. On the N. Y. municipal 
ferry boat "Richmond," one hundred 60-watt 
high-efficiency carbon lamps were replaced 

watts per sq. ft., while the latter gave 3.36 
foot-candles for 1.50 watts per sq. ft. 
Figs. 5 and 6 show this cabin before and 
after the change. 

Battleships present a different lighting 
problem from any other class of vessels. A 
warship's ability to fight, fight hard and 
effectively, is its primary function. The 
lighting must be so arranged as to enable the 
men to use the apparatus to the best advan- 
tage. A battleship is also the business office, 
the factory, the recreation ground, and the 
home of a thousand men. The happier and 
more contented these men are, the more 
efficient will they be. Realizing that plenty of 
light correctly applied increases the efficiency 
of the human machine, the U. S. Government 
has taken up the lighting of its ships from a 
scientific standpoint. Drawn-wire tungsten- 
filament lamps have proved by actual test 
that they will stand up under the severe 
strains of battle practice, the tropics, the 
extreme cold of a Maine winter, a storm, and 
a full-power run; in short all the various con- 
ditions encountered by our ships. These 
lamps in connection with scientifically de- 
signed reflectors are now being generally 
adopted by the Navy for use aboard ship. 

On ocean steamers, great attention is paid 
to obtaining sesthetic effects, as illustrated by 



the photograph of the palm garden of the 
Hamburg- American Line steamer "Victoria 
Louise," Fig. 7. Investigation shows that 
round-bulb all-frosted tungsten-filament 
lamps are most generally in use for this class 
of ship lighting. Overhead lighting supple- 
mented by table lamps and lamps in wall 
brackets is found to be a most usual arrange- 
ment. The illumination intensities vary 
quite widely, averaging however about 2 
foot-candles on a plane 3 feet above the floor. 

The introduction of the rugged drawn-wire 
tungsten-filament lamp has opened up a big 
field for its application to practically all 
classes of marine lighting. 

As the result of many tests, several steam- 
ship companies are arranging to equip their 
boats with drawn- wire tungsten-filament 
lamps in place of the carbon lamps previously 

In many instances it is impractical to 
change the existing wiring of a ship and, for 
this reason, the recommendations for a new 

ship would vary considerably from those for 
a ship at present in commission. When the 
wiring is already installed, the expense in- 
volved in changing the outlets may more 
than offset the advantages to be secured bv 
such a rearrangement of units as will provide 
the most economical and effective operation. 
On the ships which have been studied it was 
usually apparent that a better economy, and 
often a more effective illuminating effect 
could be produced by the use of a smaller 
number of tungsten-filament lamps of higher 
c-p., but in every case it was considered 
unwise to change the existing outlets. How- 
ever, advantage has been taken of the higher 
efficiency of the tungsten-filament lamp to 
increase the intensity and, at the same time, 
reduce the lighting cost by substituting, lamp 
for lamp, 25- and 40-watt tungsten-filar" 
lamps for the high wattage carbons. I 
instances it was possible to also impro^ 
diffusion by substituting frosted bulbs 
clear ones. 



Part V. (Nos. 29 to 31 Inc.) 

By E. C. Parhaai 

Construction Department, General Electric Company 

Neglecting the no-load or magnetizing 
current of a constant-potential transformer, 
the primary current is proportional to the 
secondary current as is shown by the follow- 
ing succession of actions. Increased second- 
ary current increases the opposing secondary 
flux which neutralizes more of the primary 
flux and thus decreases the counter e.m.f. of 
the primary coil, thereby permitting the 
primary current to increase until the core 
flux is restored to approximately its no-load 
value. In other words, the core flux is main- 
tained at a practically constant value because 
the primary current automatically responds 
to changes in the secondary current. 

With current transformers the conditions 
are different. The value of the primary 
current depends upon an external load over 
which the current transformer has no con- 
trol. Assume that the current-transformer 
primary current, which is the current out- 
put of some generator, is kept constant at 
full-load value. A suitably calibrated instru- 
ment, placed in the secondary circuit, will 
indicate the value of the primary current, 

although the secondary current to which the 
indication is due is but a small known part 
of the primary current. Owing to there being 
many more secondary than primary turns, 
the ampere-turns of the two windings are 
approximately equal and the opposing flux 
caused by the secondary is able to neutralize 
the flux of the primary to such an extent as to 
keep the core flux density below the value 
that would generate objectionable heating. 

Next assume the secondary to be short- 
circuited, another instrument placed in series 
with the existing one, and the temporary 
short-circuit (used to avoid opening the 
secondary) removed. The secondary resist- 
ance will now be greater than it was with but 
one instrument in circuit and consequently 
there will be a tendency for the secondary 
current to decrease. Any decrease in the 
secondary current, however, increases the 
core flux because less of the primary flux will 
be neutralized. The increased flux cutting 
the secondary turns increases the secondary 
e.m.f. and thereby restores the secondary 
current almost to its former value. Up to 
the instrument capacity of the transformer. 



additional instruments further increase the 
core flux and, hence, the secondary e.m.f. 
The secondary current is thus maintained 
so close to the correct value that the error 
may be neglected except in precision work. 

The greater the number of instruments 
used in series, the greater will be this error, 
the hotter the transformer will become on 
account of the greater flux density at which 
the iron is worked, and greater will be the 
voltage to which the secondary insulation 
is subjected. 

If the secondary be opened, its flux-neu- 
tralizing capacity ceases entirely and the 
core flux density, due to the unopposed pri- 
mary flux, becomes abnormal and the heat 
resulting from rapidly reversing the flux at 
this density soon raises the iron to a tem- 
perature dangerous to the coil insulation. 
Furthermore, this greatly increased flux 
cutting the secondary turns induces therein 
an e.m.f. far exceeding that which obtains 
under normal conditions. These two effects 
conspire to break down the secondary in- 

Therefore it is necessary to issue a warning 
against operating a current transformer with 
its secondary circuit open. 


Repulsion-induction motors operating 
under normal conditions are characteris- 
tically free from sparking. Even when ir- 
regularities exist, the sparking may be so 
slight as to mislead one who is accustomed 
to operating only the other kinds of com- 
mutator motors. 

Once, an operator noticed that one of his 
press motors was overheating and was spark- 
ing; similar motors operating similar presses 
were giving no trouble. In an effort to 
determine the cause of the difficulty, the 
operator interchanged two of the motors — 
a good one and the one under consideration. 
The trouble remained with the same motor. 
This cleared the press and the starter from 
suspicion and focused his attention upon the 
motor. He then cut an ammeter successively 
into the stator circuit of each of three dupli- 
cate press motors, which had evidenced no 
trouble, and thereby found that the current 
required under regular working conditions 
was 2J/2 amperes. The faulty motor was 
operating under exactly the same conditions 
of load as were the three good motors just 
mentioned, but the ammeter when placed in 
its circuit indicated 5 amperes. Since the 
full-load current rating of the motors was 
3.8 amperes the motor under examination 

was carrying a current overload of nearly 30 
per cent. 

Reversing the compensating field con- 
nections only made matters worse. By 
shifting the brushes a little at a time, he found 
a position where the current was of the same 
value as for the good motors. He was now 
reasonably sure that the motor would not 
overheat, but as the sparking had not been 
lessened he called in a repair man whose 
examination disclosed an open-circuited arma- 
ture lead. This condition was indicated by 
the burning out of the mica between dia- 
metrically opposite pairs of commutator bars. 
The repair of this lead resulted in perfectly 
normal operation. 


The amount of water delivered by a cen- 
trifugal pump depends, among other things, 
upon the speed of the impeller. The delivery 
and speed variations, however, are not 
directly proportional to each other for, by 
reason of the characteristics of this type of 
pump, a certain per cent increase in speed 
produces a greater per cent increase in the 
amount of the water delivered. Therefore, 
if on account of high or low line voltage, or on 
account of design irregularities, the speed of 
the motor of a motor-driven centrifugal 
pump falls below or rises above the per- 
missible five or six per cent speed variation, 
the water delivery may be materially affected ; 
and, in the case of increased speed, the motor 
may be overloaded seriously. All conditions 
will be covered in a specification stating the 
speed at which the pump shall deliver water 
at a certain rate without the motor tempera- 
ture rise exceeding a certain number of 
degrees after the motor has operated con- 
tinuously for a stated length of time. 

A certain operator once complained that 
his pump motor was running "red hot" on 
regular duty but was otherwise entirely 
satisfactory. Ammeters connected into the 
motor circuit showed the motor was heavily 
overloaded. Measurement of the water 
delivery rate showed it to be about twice 
that which had been specified. The pump 
maker was notified and his subsequent 
investigation revealed the fact that the wrong 
impeller had been supplied with the outfit. 
The pump manufacturer got out of his diffi- 
culty to advantage, however, because the 
operator was so much pleased with the water 
output that he had been obtaining that he 
readily agreed to pay the difference in price 
between the motor that had been furnished 
and the one that was large enough for the 




Standardization Rules 

The January issue of the Proceedings of 
the A.I.E.E. contains an article by Dr. A. E. 
Kennelly. Chairman of the Standards Com- 
mittee, of which the following is an abstract: 

The American Institute of Electrical Engineers 
maintains, among its standing committees, a 
"Standards Committee,'' which is charged with the 
important duty of maintaining a series of Standard- 
ization Rules for the benefit of the Institute and its 
members. The fifth and most recent edition of the 
Standardization Rules, which is dated December 1, 
1914. was adopted by the Board in July, 1914. It 
covers 96 pages. 

Object of the Rules. The main purpose hitherto 
aimed at by the Standards Committee in the rules 
has been to draw up engineering definitions of terms, 
phrases, and requirements, relating to electrical 
machinery and apparatus, so that the meaning of 
technical terms might be standardized among the 
members of the Institute. Particular effort has been 
directed towards defining, in engineering terms, the 
rating of electrical machinery, and the requirements 
connoted thereby. This work serves the entire 
electrical industry, including manufacturers, pur- 
chasers, technical advisers, operators, and con- 
sumers. It is therefore desirable that the represen- 
tation of electrical interests on the Standards Com- 
mittee should be as wide as possible, in order that 
the needs of all classes of electrical workers should 
be adequately presented and mutually protected. 

Relations of the Standards Committee to other 
Engineering Bodies. The work of the Standards 
Committee naturally brings the committee into 
contact with bodies engaged upon standardization 
in neighboring fields. Thus, it has for a number of 
years co-operated with the Bureau of Standards at 
Washington, D. C. The Bureau has not only been 
continuously represented by one or more of its 
i .fixers on the Committee, but it has also under- 
taken important researches in electrical engineering 
at the request of the Committee. Thus, in 1910, 
the Bureau, at the request of committee, made an 
extensive investigation into the conductivity of 
commercial copper and has since published complete 
copper wire tables* based on those researches, for 
the benefit of the electrical industry. 

The A.I.E.E. Standards Committee has also at 
different times worked in co-operation with Stand- 
ards Committees of the American Society of Me- 
chanical Engineers, the Illuminating Engineering 
Society, the Institute of Radio Engineers, the 
American Society for Testing Materials, the National 
Electric Light Association, the Association of Edison 
Illuminating Companies, and other bodies, upon 
questions of standardization involving work in their 
ive fields. It is probable that a standards 
committee representing the entire engineering force 
of America may ultimately be secured for dealing 
with general engineering questions. 

International Standardization. The conditions 
which affect the mutual relations of different engi- 
neering societies in America regarding standardiza- 
naturally extend themselves internationally 
rresponding engineering societies in other 

t. *_ Cir< ?T lar - N '°- 31 of the Bureau of Standards, "Copper Wire 

countries. It becomes impossible to carry stand- 
ardization beyond a very elementary stage in any 
one country, without influencing the work and pro- 
cedure along similar technical lines in other count :ies. 
It therefore becomes desirable to enter into mutually 
co-operative relations with electrical engineering 
societies and their standardizing committees 
abroad. Co-operative relations have been entered 
into at different times between the A.I.E.E. Stand- 
ards Committee and corresponding committees 
in other countries, to considerable mutual advan- 
tage; but especially through the influence of the 
International Electrotechnical Commission, an 
international body engaged in international electrical 
engineering standardization. The American Stand- 
ards Committee of the A.I.E.E. was the first 
national committee to formulate and publish elec- 
trical standardization rules, and similar com- 
mittees have since come into existence in various 
other countries. It is neither necessary nor desirable 
that electrical apparatus built in one country should 
conform in structural details to that built in other 
countries; but it is surely desirable that the rating, 
and rating terms, employed in specifying the be- 
havior in different countries should correspond, 
since no country can permanently profit by ambi- 
guity, in the meaning of its technical phraseology, 
as applied to the physical behavior of apparatus. 

The Standards Committee of the A.I.E.E. has 
no direct representation on the International Elec- 
trotechnical Commission (I. E. C); but it has close 
relations with the U. S. National Committee of the 
I. E. C, and, through the intervention of the latter 
committee, it has been able to present its needs and 
recommendations to the I. E. C. Various rulings 
of the I. E. C. at past international meetings are 
now incorporated in the latest edition of the Rules. 

The Relations of the Standards Committee to the 
Institute Membership. The work of its Standards 
Committee constitutes a distinct asset to the 
Institute, and to the membership. The committee 
meetings usually occur at monthly intervals in the 
New York headquarters of the Institute. Notices 
of these meetings are communicated in advance to 
the Standards Committees of other engineering 
societies, and are regularly announced at head- 
quarters. Suggestions regularly reach either the 
A.I.E.E. Secretary, or the Secretary of the Com- 
mittee, and receive careful consideration by the 
Committee. It is very desirable to secure frequent 
and copious suggestions, from the Institute mem- 
bership at large, as to how the Rules operate in 
practice, and how they may be improved. 

The Standards Committee is thus a body earnestly 
devoting its time and service to the welfare of the 
electrical engineering industry, in the belief that 
precision in standardization means an advance in 
the ethics, the science, the business and the welfare 
of engineering. 

The 303rd meeting of the American Insti- 
tute of Electrical Engineers was held at the 
Engineering Societies Building in New York, 
on Friday, January S, at 8:15 p.m. 

Mr. I. W. Gross presented a paper 
entitled, Theoretical Investigation of Electric 
Transmission Systems Under Short Circuit 
Conditions. The leading features of the trans- 



mission system under short circuit conditions 
were treated as follows : 

First. Mechanical forces between the 
phases of three-conductor, three-phase cables 
when carrying short-circuit current. Under 
this heading the forces between busbars were 
also investigated. 

Second. The heating of the conductors of 
the cable from the instant of short circuit to 
a time 0.8 second later was traced analyti- 
cally, during the transient state of the current, 
and typical computed heating curves were 

Third. The effectiveness of the method of 
placing reactors between generator terminals 
and the bus from which power is taken, and 
additional reactors between generators and an 
auxiliary synchronizing bus were analyzed. 
This latter scheme was compared with the 
present well-recognized schemes of feeder and 
busbar reactors. 

It was shown that the average mechanical 
forces existing between conductors of a three- 
phase cable carrying short-circuit currents. 
and between busbars, rise to relatively high 
values at the instant of trouble. These forces 
can be reduced in cables by either limiting 
the current or increasing the distance between 
conductors. The same applies to busbars, 
and in addition, the position of the busbars 
can be adjusted by placing them in the same 
plane so that the mechanical forces may be 
considerably reduced. 

The heating of the cables may be the limiting 
feature in controlling the short-circuit current, 
since it is quite possible for the temperature 
of the conductor to rise to such a point as to 
endanger the insulation of the cable even in 
the very short time that it takes an oil switch 
to operate after the short circuit has occurred. 
When the characteristics of the generators 
under short-circuit conditions are known, it is 
possible to compute the temperature rise, 
even although the current is of transient 

In using reactors to limit the current flow 
on a power system, the method of plain 
feeder reactance is not fully effective, as 
trouble on the main station bus is almost 
certain to cause to drop out of step all syn- 
chronous apparatus on the system. Further, 
this method offers no protection to machines 
against poor synchronizing. 

Station busbar reactance is effective under 
short-circuit conditions, but under normal 
operation is objectionable on account of 
the large voltage drop in transmitting power 
from one end of the bus to the other. 

The scheme of feeding from the machine 
terminals, and paralleling generators on a 
separate bus, as brought to light by Mr. 
Stott, is extremely flexible and very effective 
in furnishing protection. It can limit the 
current to a safe value without an excessive 
amount of reactance in the circuit; it can 
protect the machines against mechanical 
injury due to poor synchronizing; and can 
transmit power from between different points 
of the bus with far less voltage drop than with 
the bus reactance scheme. It makes possible 
the use of generators having a low inherent 
reactance, provided, of course, the machine 
is designed to withstand dead short circuit 
at its terminals. Further, the lower the 
reactance of the generator the less is the 
probability that the synchronous apparatus 
on the system will be out of step due to 
reduced power house voltage. The possibili- 
ties of this system are as yet probably not 
fully realized. 

The full text of the paper appears in the 
January issue of the Proceedings of the 


The regular meeting of the Lynn Section 
of the A.I.E.E. was held on the evening of 
January 6th and was attended by 250 mem- 
bers. Prof. Comstock, of the Massachusetts 
Institute of Technology, gave a most interest- 
ing talk on the Modern Theory of Electricity 
and Matter. The subject was introduced by 
a review of the older theories of the molecular 
constitution of matter and the kinetic theory 
of gases, and it was pointed out how the recent 
measurements of the Brownian movements 
of particles confirmed the kinetic theory, and 
how in the spinthrascope the effect of the 
impact of single particles of atomic magnitude 
is made evident. The general magnitude 
of molecules and atoms was described. 

Electrical discharges through gases were 
discussed and it was pointed out that here we 
have the simplest means of studying the 
phenomena of electric currents. The vacuum 
discharge consists in the actual transport of 
particles of electricity, and the particles 
involved are all of one size irrespective of the 
material from which they are torn, and all 
carry the same electric charge, which is the 
smallest quantity of electricity known to 
exist, and is called the "Electron." Further, 
these particles have a mass of about 1/2000 
part of that of the hydrogen atom. To 
illustrate the intensity of the electrical forces, 
it was stated that could two particles of the 



size of pin-heads, located a mile "apart, carry 
charges in proportion to that of the electron, 
they would act upon each other with a force 
of hundreds of tons. 

The electric current in a wire was described 
as a slow migration of electrons and the con- 
trast of the speed of the electrons to the speed 
of propagation of electric disturbances was 

The talk was illustrated by a number of 
lantern slides and by many apt analogies. 
The statements relative to the numerical 
magnitudes of the quantities discussed made 
the talk most interesting and instructive. 

On February 3rd, Major J. A. Shipton will 
speak before the Lynn Section on a military 
topic, and on February 17th, Mr. J. L. Wood- 
bridge will speak on The Characteristics and 
Uses of Storage Batteries. 


On the 7th of January Prof. W. S. Frank- 
lin, of Lehigh University, read a paper to the 
Section on the subject of Electric Waves. His 
paper was illustrated with lantern slides and 
dealt mainly with the derivation of the 
fundamental equations of wave motion. 
It is hoped that an abstract of this lecture 
may be given in a later number of the 

On Friday, January 29th, Dr. Irving 
Langmuir of the Research Laboratory, 
Schenectady, will lecture to the Section on 
the subject of Modern Theories of Electricity. 


There was a meeting of the Schenectady 
Section of the A.I.E.E. at Edison Club Hall, 
on the 5th of January. The general subject 
for the evening was Electric Illumination and 
after a few introductory remarks by Mr. 
S. H. Blake, the following papers were 
presented : 

The Arc as a Street Ilhtminant, by Mr. 
(". A. B. Halvorson, Designing Engineer 
of the Arc Lamp Department at the Lynn 

nparison of the Operation of Low-Current 
and High-Current Gas-Filled Lamps on Scries 
Circuits, by H. D. Brown of the Consulting 
Engineering Department. 

I haracteristics of Gas-Filled Mazda Lamps, 
by Mr. L. A. Hawkins. Research Laboratory. 

The papers were illustrated by lantern 
slides and by means of a large variety of 
experimental apparatus in operation. 

Mr. Halvorson first brought out the point 
that to render an illuminant suitable for street 
lighting sen-ice even a very simple lighting 
unit like a high-current mazda lamp must be 
provided with various attachments such as 
a weather protective casing, a compensator 
coil, socket, globe holder, refractor or outer 
globe, etc.. which in the end makes the 
complete outfit look very much like the 
familiar street-lighting arc lamp. 

He then showed a vertical distribution curve 
of the light from a clear-glass lamp of the gas- 
filled mazda type, without reflector, refractor 
or diffusing globe. This curve indicated that 
as much light is thrown above the horizontal 
as below. For street illumination the most 
useful light is that which is distributed on 
an average of 10 degrees below the horizontal. 
In order therefore to utilize as much of the 
total light flux of the lamp as possible, it is 
necessary to equip the lamp with a suitable 
reflector and refractor which redirect the 
light from the upper hemisphere and from 
under the lamp. By this means the maximum 
light is thrown in the desired direction. Similar 
curves were then shown of the light distribu- 
tion of the flame lamp, the titanium lamp, and 
of the magnetite lamp with high-efficiency 
electrodes, with and without redirecting 
and diffusing devices. Samples of the various 
lamps mentioned were shown in operation. 

Mr. Halvorson then analyzed and de- 
scribed the evolution of the refractor, the 
recent introduction of which into practical 
use in connection with large street lighting 
units he regarded to be of great importance. 
He showed that without the refractor a 
single reflector of suitable shape 16 feet in 
diameter would be necessary to redirect all 
the upward light from an arc lamp. By the 
use of biplane or triplane small-diameter 
(about 20 inches) reflectors it is possible to 
obtain the equivalent of the effect of this 
very large reflector. However, the refractor 
has the further advantage that it not only 
redirects the upward light but also the 
light from under the lamp. The construction 
of the refractor was shown to consist of two 
clear-glass cone-shaped globes ground so that 
one fits perfectly inside of the other. On the 
outside surface of the inner globe, horizontal 
prisms arc moulded to give the light-re- 
direction desired, while on the inside of the 
outer globe are moulded vertical prisms 
designed to diffuse the light. When the 
globes are fitted together the combined unit 
presents smooth surfaces inside and out, so 
that it can readily be kept clean. 



Curves were shown of the very remarkable 
improvements in light efficiency obtained 
with magnetite lamps by the use of the new 
"high-efficiency" electrodes. It was further 
shown that by the use of flattened electrodes, 
about one-quarter inch wide, in place of round 
electrodes of the same cross-sectional area, 
not only is better distribution obtained, 
but higher efficiency, greater steadiness, 
cleaner burning, and it is possible to run the 
arc satisfactorily at lower current and arc volt- 
age. This latter fact means that the lamps 
can be put out in smaller units, a matter 
of the greatest importance in arc lighting. 

A large ornamental globe and fixture of 
the type used for street illumination in 
Washington, D. C, was shown, and in the 
globe were mounted a magnetite arc lamp 
and a high-current mazda lamp of equal 
wattage. The lamps were switched on 
separately so that the color and intensities 
of the lights could be compared. 

In closing, Mr. Halvorson commented 
briefly on the subject of maintenance and 
operating costs. 

The paper indicated that arc lamp devel- 
opments have been keeping step with the 
remarkable advances that have been made in 
the incandescent lamp field and show that 
the arc lamp, for street lighting purposes at 
least, is a very important factor. 

In Mr. Brown's paper the author explained 
that with the development of gas-filled lamps 
there appeared two types, designated re- 
spectively as "low-current" and "high-cur- 
rent." He compared the use of these two 
types on similar series circuits. The low- 
current type may be operated directly on 
existing series systems, while the high-current 
type is adapted to standard circuits by 
means of auto-transformers whose functions 
are to transform the line current to the 
proper value for the lamps, and also to 

protect the lamps against abnormal condi- 
tions. From calculations based on experi- 
mental and design data, the following 
conclusions were drawn: 

(1) For abnormal conditions of primary 
voltage fluctuation or accidental shorts on 
the secondary, the compensator units are 
better protected against extreme conditions; 
and, even for moderate fluctuations, the result- 
ing effect on the lamp is appreciably less and 
therefore the use of compensators offers a 
greater reliability. 

(2) For normal operation it is shown that 
the lowering of the primary power-factor, 
due to the . introduction of compensators 
in the secondary, is not a serious loss, since 
the use of the more efficient high-current 
lamps allows the operation of a few more 
units for full load on the series transformers 
and consequently more available light of a 
better quality. 

Mr. Hawkins in his paper very clearly 
explained the principle of operation of gas- 
filled incandescent lamps, the various stages 
of their development, and the reason why 
this construction which now seems so simple, 
was not obvious before. He carried out two 
striking experiments, one showing the very 
remarkable improvement in candle-power 
and efficiency effected simply by winding the 
tungsten-wire filament of the gas-filled lamp 
in the form of a tight spiral instead of looping 
it on supports as with vacuum lamps, and 
the other showing the marked advantage 
to be obtained from the use of high-current (15 
to 20 amp.) over low current (6 to 7)4 amp.) in 
securing high-efficiency at equal wattage. 

On the 19th of January, Dr. W. D. Coolidge, 
Assistant Director of the Research Laboratory 
of the General Electric Company, presented 
a paper entitled Recent Developments with 
X-Rays. This paper will be abstracted in 
the March issue of the Review. 





In view of the remarkable improvements in X-ray 
generators, recently achieved by Dr. W. D. Coolidge 
and Dr. I. Langmuir in the General Electric Re- 
search Laboratory, and in view also of the wonderful 
use of these rays lately made by Dr. W. H. Bragg 
and his son, W. Lawrence Bragg, in the determina- 
tion of the atomic structure of crystals, etc., it may 
be timely to present a brief retrospective sketch 
pertaining to the general subject of X-rays. 

From time almost immemorial three states of 
matter have been recognized: The solid, the liquid, 
and the gaseous; but in the year 1816, that pro- 
found philosopher, Michael Faraday, conceived of 
its existence in a fourth state, to which he gave the 
peculiar and significant name of "Radiant Matter," 
and he considered this state of matter to be as 
distinctly different from the gaseous state as the 
gaseous is from the liquid, or the liquid from the 

In 1879, 63 years after the date of Faraday's 
remarkable conception, Sir William Crookes delivered 
his memorable lecture on "Radiant Matter" before 
a meeting of the British Association at Sheffield, 
England. In this lecture he exhibited and described 
very highly exhausted tubes which were almost 
identical in outside form and internal construction 
with some of our present X-ray tubes, being fur- 
nished with large concave cathodes, small anodes, 
and intermediate pieces of iridio-platinum, the 
latter corresponding to the targets of our X-ray 

Sir William Crookes unquestionably produced 
X-rays when he made his classical experiments on 
"Radiant Matter," but he did not chance to bring 
any substance within range of their influence that 
would have made them directly or indirectly ap- 
parent, so their existence at that time remained un- 

In 1893, fourteen years after the delivery of this 
lecture, we hear of Dr. Philip Lenard experimenting 
with Crookes' tubes in Heidelberg, Germany. Len- 
ard made a tube with a little aluminum window- 
through which the cathode rays could shine out 
into the open air, whereas they had previously 
been confined to the inside of the tube by the glass ' 
wall which they could not penetrate. Lenard then 
discovered that these rays, which have been called 
"Lenard Rays," could be deflected by magnetism, 
could produce photographic action and cause fluores- 
cence in certain substances, barium-platino-cyanide 
being affected in the highest degree. 

Two years later we find Prof. Wilhelm Konrad 
Rontgen also experimenting with Crookes' tubes in 
the Institute of Physics in Wurzburg, Bavaria. 
He covered one of these tubes with black paper, and 
when it was excited by the current from an induction 
coil, noticed that a piece of cardboard which had 
been coated with barium-platino-cyanide, and was 
lying on the table near by, glowed with a bright 
green , e , although the black paper com- 

every ray of visible light that was 
produced in side the tube bv the electrical discharge. 

Rontgen's trained intell him at onci 

arkable phenomenon with the exist- 

o unknown radiation, to which 

both the glass of the tube and its covering of black 

*S« Dr. Dence Jones' "Life and Letters f Faraday," Vol. I. 

paper were transparent, and which caused the 
fluorescence of the barium-platino salt. Placing his 
hand between the glowing screen and the darkened 
tube he saw not only the dim shadow of his hand on 
the shining surface, but also the darker outlines of 
the bones within. 

Thus were the X-rays at last brought to light: 
foreshadowed in Faraday's conception of "Radiant 
Matter" in 1816: actually produced, but unrecog- 
nized, by Crookes in 1879; carried to the very verge 
of revelation by Lenard in 1893; and discovered by 
Rontgen on November 8, 1895. 

The announcement of his discovery was made by 
Dr. Rontgen in a paper read at the Institute of 
Physics of the University of Wurzburg in Bavaria, 
in December, 1895. 

In this paper the method of generating X-rays 
was explained, their fluorescent effect on a cardboard 
screen coated with barium-platino-cyanide was 
described, and their action on the photographic 
plate recorded as a fact of special significance. The 
relative transparency of various bodies to the newly 
discovered rays was also noted, and some negative 
results were mentioned regarding attempts to reflect 
and refract them. He also stated in this paper 
that the term "rays" was used for the sake of 
brevity, the prefix "X" being given to distinguish 
them from other rays, such as Lenard's for example. 

On January 7, 1896, the news of Prof. Rontgen's 
marvelous discovery was cabled to this country and 
its extraordinary character and value were imme- 
diately recognized, while its novel and almost magical 
possibilities appealed so strongly to the public at 
large that the announcement spread in an incredibly 
short period. 

Some time naturally elapsed after the discovery 
of X-rays by Dr. Rontgen before their curative as 
well as their destructive qualities were fully recog- 
nized. During this period many ardent and enthu- 
siastic experimenters received serious injuries as as 
result of their inexperience, and in a few lamentable 
cases X-ray burns produced even fatal results. 

This stage of unfortunate ignorance, however, 
soon passed, and improvements in X-ray apparatus 
generally were rapidly developed, so that the rays 
could be properly administered without danger to 
either the patient or the operator, and at the present 
time every well equipped hospital in the country 
has an efficient X-ray generating apparatus. Thus 
in the course of less than twenty years Dr. Rontgen's 
great discovery has developed into one of the most 
beneficent agents for the alleviation of many bodily 
ailments, and by reason of the later improvements, 
referred to in the beginning of this article, their 
further scope, not only in medical treatment, but 
also in many branches of scientific research, will 
be greatly enlarged. W. S. Andrews. 

ERRA TA: Attention is called to two corrections to 
apply to the article "From the'Consulting Engineering 
Department of the General Electric Company" in the 
January number of the GEXERAL ELECTRIC 

Twenty-one lines from the bottom of the right-hand 
column, p. 73, the equation as printed 

,' Edi /* /•■ Edi /*• 

I (a + bi + 'ciVi I dt should be \ 'a+bi +«•) i I dt 

Eight lines from the bottom of the same column the 
word "infinite" should be "infinite". 

General Electric Review 


Manager. M. P. RICE Editor, JOHN R. HEWETT Associate Editor, B. M. EOFF 

Assistant Editor, E. C. SANDERS 

Subscription Rates: United States and Mexico, $2.00 per year; Canada, $2.25 per year; Foreign, $2.50 per year; payable in 
advance. Remit by post-office or express money orders, bank checks or drafts, made payable to the General Electric Review 
Schenectady, N. Y. 

Entered as second-class matter, March 26, 1912, at the post-office at Schenectady, N. Y., under the Act of March, 1879. 

VOL. XVIII., No. :i by c^T&WLtany March, 1915 


Supplement: Reproduction in Colors of Resolutions Presented to C. A. Coffin and 
E. W. Rice, Jr., by the Association of Edison Illuminating Companies 

Frontispiece . I."i4 

Editorial: The Paths of Progress 155 

A New Device for Rectifying High Tension Alternating Currents. . . 156 

By Dr. Saul Dushman 

Parallel Operation of Alternating Current Generators Driven by Internal Combustion 

Engines 167 

Part I : Factors Affecting Generator Design. 

By R. E. Doherty 

Part II: Factors Affecting Engine Design. 

By H. C. Lehn 

Tests of Large Steam Hoists . 179 

By H. E. Spring 

High Voltage Arrester for Telephone Lines .... . 1S9 

By E. P. Peck 

X-Ray Examination of Built-Up Mica . 195 

By C. N. Moore 

The Effect of Chemical Composition upon the Magnetic Properties of Steels . 197 

By W. E. Ruder 

Electrophysics: Electron Theory of Electric Conduction in Metals . . 204 

By J. P. Minton 

Lock Entrance Caisson for the Panama Canal . . .210 

By L. A. Mason 

Practical Experience in the Operation of Electrical Machinery, Part VI . 217 

Excessive Contact-Shoe Pressure; Electric Brake Adjustments; Rotor Rubbed 
Stator; Jerky Motor Acceleration. 

By E. C. Parham 

A Hydro-electric Installation on a Coffee Plantation . . 219 

By J. H. Torrens 

Notes on the Activities of the A. I. E. E. . • 222 

From the Consulting Engineering Department of the General Electric Company . 226 

Question and Answer Section . . • 2-7 

























































































































|— > 

























With this issue we publish a unique 
supplement. It gives us special pleasure to 
meet the request to give greater publicity 
to the resolutions which we reproduce, as 
we feel that from the very nature of things 
few are in a position to recognize the 
great service done the electrical industry by 
some of those who were so active in its 

Twenty-five years ago there were some 
few active workers who by their energy, 
faith and forethought were laying the founda- 
tion stones on which the electrical industry 
of today stands. The number of workers 
soon greatly increased and some, that un- 
doubtedly did much to further the great 
enterprise, have been lost sight of, but it is 
inevitable in all human affairs that many 
workers never get their due reward. 

The electrical industry of today when 
viewed from its broadest aspect is the' 
greatest of all industries and has, so far as 
the human mind can judge, the greatest 
future before it. No one factor with the 
single exception of the invention of the 
steam engine, with which it is so inseparably 
connected, has done more to bring about 
the changes in our economic and social mode 
of living which we have witnessed during 
the last quarter of a century. When we add 
to the electrical industry proper the activities 
it has stimulated in a host of others such as 
the iron, copper, power, lighting and railway 
industries, etc., we are forced to a recognition 
of the growth of the electrical industry as 
the most potent factor in modern industrial 
life; and when we add to these the scientific 
accomplishments which would have been 
non-existent but for the advent of an electrical 
age we have to acknowledge that the electrical 
industry has been a most mighty factor in our 
modern intensive scheme of civilization. 

In some future issue we shall attempt to 
review the electrical industries and show 
their scope, but what we are particularly 
interested in at this present writing is the 
thought — whence all these wonderful develop- 
ments have sprung. 

A final analysis could only lead to the 
conclusion that it is a tremendous triumph 
of mind over matter — the useful forces of 
nature converted to the service of man. 
It is of special importance to note that these 
great things have been accomplished by the 
mind of man firstly, by the hand of man 
secondly; the thought came before the work 
and accomplishment, and that during the last 
quarter of a century we have been witnessing 
to a greater extent than ever before the de- 
velopment of the experiments of yesterday, 
which are the first fruits of productive 
thought, into the industries of today. 

The men to whom the electrical industry 
owes most today are those who thought of 
its possibilities and the part it could play 
in our future development. Without their 
activities and forethought, without their 
resourcefulness and faith, the work of a great 
industrial army would never have been 
brought into being. The type of mind that 
has the power to originate and the type of 
mind that has the power to organize are the 
greatest capital assets that the industrial 
world possesses. These are the foundation 
stones on which the whole fabric of our 
modern structure rests. In spite of anarchy, 
in spite of socialism, in spite of government 
and mis-government, in spite of the ever 
continuing war of the many against the few. 
of those that have against those that have 
not, the work of the brain must always take a 
higher stand, whether the thinkers get their 
reward or not, than the work of the hand. 
Both types of the work are absolutely 
essential to our modern scheme of life, but 
it is inevitable that the creative genius must 
always be greater than the hand which fash- 
ions the material thing created by the mind. 

The foundations of the great operating 
companies were being laid at the same time as 
those of the great manufacturing companies, 
and so it gives us special pleasure to record a 
tribute from the representatives of the great 
Edison Illuminating Companies to two of those 
who have not only been pioneers of the in- 
dustry, but who have been active workers 
from its inception up to the present time. 





By Dr. Saul Dushman 
Research Laboratory, General Electric Company 

In the following paper the writer discusses an interesting application of the theoretical investigations 
on electron emission from incandescent metals. The construction of a high voltage rectifier illustrates the 
old expression that the theory of the present may become engineering practice of the future. — Editor. 


The emission of negatively charged cor- 
puscles or electrons from heated metals may 
be illustrated by the following arrangement. 
In an ordinary lamp bulb containing a 
tungsten or carbon filament there is also 
sealed in a metal plate. After the lamp is 
well exhausted it is observed, on charging 
the filament negatively (making it cathode) 
with respect to the plate, that a current 
passes across the vacuous space. If the 
filament is charged positively this current 
disappears. Furthermore, the magnitude of 
this electron emission (thermionic current) 
from the heated cathode increases with 
increase in the temperature of the filament. 

This effect had been observed by Edison 
and was more fully investigated in the case 
of carbon lamps by Fleming.* In view of the 
unilateral conductivity possessed by such 
an arrangement as that described above, 
Fleming applied it as an "electric valve" to 
rectify electric oscillations such as are ob- 
tained from a "wireless" antenna, and 
therefore render it possible for these oscil- 
lations to affect a galvanometer or telephone. f 

That the current from a hot cathode in an 
exhausted bulb is due to a convection of 
electrons, that is of negatively charged 
corpuscles having a mass which is about 
1 1 MMJth of that of a hydrogen atom, may be 
shown by deflecting the current in magnetic 
and electrostatic fields and determining the 
ratio e/tn. Another method is that described 
below and which depends upon the space 
charge produced by the electrons under 
certain conditions. 

The relation between thermionic current 
and temperature of cathode was further 
investigated by Richardson and he found 
that in all cases the relation could be accu- 

•Proc. Roy. Soc.. Lond.. J,7, 122 (1890). 

tProc. Roy. Soc.. Lond.. 74. 476 (1905). See also J. A. 
tleming, Principles of Electric Wave Telegraphy and Telephony 
pp. 477-482 (second edition). * 

rately represented by an equation of the form, 


i = ay/jt (i) 

where a and b are constants for the particular 
metal and i is the saturation thermionic 
current per unit area at the absolute tem- 
perature T. 

Subsequent experiments, however, by other 
investigators tended to throw much doubt 
upon the actual existence of a pure electron 
emission from a heated metal in a good 
vacuum. It was found that different gases 
affected the values of the constants a and b 
to an immense extent, so that at the same 
temperature the thermionic currents obtained 
varied over a very wide range. Furthermore, 
it seemed that the greater the precaution 
taken to attain high vacuum, the smaller the 
thermionic currents obtained, and the con- 
clusion was drawn that in a "perfect" 
vacuum the thermionic currents would dis- 
appear altogether. In fact, the view generally 
held until the past year by the German 
physicists and by quite a few English phy- 
sicists was that the thermionic currents were 
due to chemical reactions in a gas layer at the 
surface of the heated metal, and that there- 
fore there was no justification for believing 
in the existence of a pure electron emission 
per ipse from a heated metal. 

This subject was taken up in the Research 
Laboratory of the General Electric Company, 
by Dr. Irving Langmuir, and he found that 
in the case of heated tungsten filaments the 
electron emission at constant temperature 
increased as the vacuum improved until a 
constant value was attained which varied 
with the temperature in accordance with 
Richardson's equation. Dr. W. D. Coolidge 
applied this fact to the construction of a hot 
cathode Rontgen ray tubej in which elec- 
trons are produced from a heated filament 
in a highly exhausted bulb. A tungsten 

}W. D. Coolidge. Phys. Rev.. Dec. 1913. 


target is used as anode and by applying very 
high voltages (50,000 to 100,000) the electrons 
are given velocities great enough to produce 
very penetrating X-rays when they strike 
the target. 

During the last few years Dr. Langmuir 
has carried out a detailed investigation of the 
whole subject of electron emission from 
heated metals and the results obtained have 
led to a large number of interesting and 
highly important applications. 

While a complete summary of these appli- 
cations will be presented by Dr. Langmuir 
at a future meeting of the Institute of Radio 
Engineers, it has been considered advisable 
to publish a preliminary account of one 
important application of hot cathode tubes 
in the development of which the writer has 
been interested. This concerns the appli- 
cation to the rectification of high tension 
alternating currents. 

The Hot Cathode Rectifier 

As mentioned above, the fact that an 
exhausted tube, containing two electrodes, 
one of which is heated by some external 
source, acts as a rectifier, has been known for 
a number of years. But difficulties were met 
with in the way of applying this practically. 
The magnitude of the current obtained was 
apt to vary quite erratically, especially with 
slight variations in degree of vacuum. Fur- 
thermore, in the types of hot cathode rectifiers 
exhausted by ordinary methods, the electron 
emission is accompanied by a blue glow. 
This glow becomes more and more pronounced 
the higher the voltage at which the rectifier is 
operated, and it is found that under these con- 
ditions the cathode gradually disintegrates so 
that the rectifier becomes inoperative. 

An explanation of these phenomena grad- 
ually developed as a result of the above 
mentioned investigations on thermionic cur- 
rents in high vacua. It was perceived that 
the blue glow is due to the presence of pos- 
itively charged gas molecules (ions), and that 
the disintegration of the cathode is due to 
bombardment by these positive ions moving 
with high velocity. But when the vacuum is 
made as perfect as possible, the conduction 
occurs only by means of electrons emitted 
from the hot cathode, and there is no evidence 
whatever of any blue glow or other forms of 
gaseous discharge. Thus, while it had 
previously been considered that a certain 
amount of gas is absolutely essential to 
obtain conduction from a hot cathode, and 
the presence of blue glow was taken to be a 

necessary accompaniment of conduction in 
such cases, it was found that by adopting 
certain methods of treatment and the use of 
high vacua, a hot cathode rectifier could be 
constructed in which all of the difficulties 
discussed above are avoided. 

Special methods have been developed for 
treating all metal parts and glass walls so 
that they are made as free of gas as possible. 
A Gaede molecular pump in series with two 
other pumps is used to evacuate the tubes. 
It has been shown by the writer* that by 
using this arrangement together with a liquid 
air trap inserted between rectifier and molecu- 
lar pump, it is possible to attain a vacuum as 
high as 5X10~ 7 mm. of mercury. At this 
pressure the mean free path of an electron 
is so great that the chance of its colliding with 
any gas molecules and thus forming ions by 
collision is reduced to a minimum. 

In the Coolidge X-ray tube there is no 
difficulty in obtaining such a good vacuum 
that no gaseous discharge occurs even when 
150,000 volts is applied across the electrodes. 
There appears to be no limit to the voltage 
for which the tube may be constructed 
except that due to electrostatic strains. A 
further discussion of this point is, however, 
reserved for a subsequent section. 

Electron Emission in High Vacuum 

Regarding the difficulty of obtaining con- 
stant values for the thermionic currents at 
given temperatures, it has already been 
mentioned that in a sufficiently good vacuum 
the results obtained are perfectly definite and 
reproducible. In the case of tungsten in a 
"perfect" vacuum the value of the constants 
a and b in the Richardson equation are 23.6 X 
10 9 and 52500 respectively, where i is meas- 
ured in milli-amperes per square centimeter-! 

Using these constants, the values of i 
calculated for different values of T are as 
given in the following table: 





4.2 milli-amps. 













*S. Dushman. Phys. Rev., April, 1914. The complete paper 
will appear very shortly in the same journal. 

fl. Langmuir, Physikal. Zeit., IS, 516 (1914). 



In Fig. 1 these results have been plotted 
on semi-logarithmic paper. Plotting directly 
values of i against those of T one obtains a 
curve of the form shown in Fig. 2.* 

Space Charge Effect 

It was observed by Langmuir that in addi- 
tion to this temperature limitation the elec- 
tron current may be also limited by space 
charge. With a low potential difference 
between the electrodes the phenomena ob- 
served are as follows: 

As the temperature of the cathode increases, 
the electron emission increases at first in 
accordance with the equation of Richardson. 
However, above a certain temperature this 
current becomes constant; further increase 
in temperature does not cause any correspond- 
ing increase in thermionic current. The 
temperature at which this limitation occurs 
.increases with increase in anode potential. 
The curves shown in Fig. 2 illustrate this very 
well. They represent the results observed 
when the thermionic current was measured 
from a 10-mil tungsten filament situated 
along the axis of a cylindrical anode 7.62 
cm. long and 1.27 cm. in radius. Thus, 
with a potential difference of 55.5 volts, the 
electron emission increased according to the 
equation of Richardson until a temperature 
of about 2300 deg. K was attained. With 
further increase in temperature, the ther- 
mionic current remained absolutely constant. 
But when the voltage was increased to 87.5, 
the thermionic current continued to increase 
up to 2350 deg. K. With a voltage of 129, 
the increase in thermionic current was 
observed up to 2400 deg. K. 

This effect (which is observed only in 
extremely good vacua) is due to the existence 
of a space charge produced by the emitted 
electrons. In other words, the electrons 
emitted from the hot cathode produce an 
electrostatic field which tends to prevent the 
motion of any more electrons toward the 
anode. As the positive potential on the latter 
increases, more and more electrons are 
permitted to reach the anode. 

From theoretical considerations it was 
deduced by Langmuir that the thermionic 
current ought to increase with the three-halves 
power of the voltage (until the saturation 

* S. Dushman. Phys. Rev.. 4. 121 (1914). The area of the 
hot filament used was 0.61 cm 3 . The experiments from which 
this curve was plotted were performed some time ago. Both 
the degree of vacuum attained and the accuracy of temperature 
determination were not as good as that obtained in measuring 
the values of. a and 6 given above. When it is considered that 
an error of 25 degrees in the determination of the temperature 
at 2400 deg. K. is sufficient to account for the difference between 
these values of the constants and those given in the curve, the 
discrepancy does not appear so great. 

current as defined by the Richardson equation 
is attained), that is, for electrodes of any 
shape, the space charge current 

*, = *. V* (2) 

where V denotes the potential difference and 
k is a constant depending on the shape of the 
electrodes, their area and the distance apart. 
For the case of a heated filament in a 
concentric cylindrical anode (infinite length) 


in r 


*' 9 \m 

where i, is the thermionic current per unit 
length and r is the radius of the anode. 

Converting into ordinary units (milli- 
amperes and volts) this equation becomes 

14 fi 

xi^xio- 3 



/ 1 
















/ Electron, Emission from "Tungsten 

1 1 

Calculated from Equation 














/ — j> 








Oeg Kelvin. 









Fig. 1. Electron Emission from Tungsten 
in a "Perfect" Vacuum 

The data shown in Fig. 2, which were 
obtained in the course of an investigation 
carried out by the writer, are in full accord 
with the results calculated from this equation. 
Substituting for r the value 1.27 cm., and 
noting that the actual length of cylinder 
used was 7.62 cm., the values of the constant 
factor as obtained from the observed space 


*>2.32X10- 3 xV 

charge currents for different voltages do not 
differ by more than 2 per cent from 14.6.* 

In the case of a heated tungsten plate 
parallel to another plate, the space charge 
current per sq. cm. 



where x is the distance between the plates 
in centimeters. 

It ought to. be observed that up to a point 
at which the diameter of the filament amounts 
to about five per cent of the diameter of the 
anode cylinder, or of the distance between 
the plates, the space charge voltage is 
independent of the actual diameter of the 

The thermionic current from a hot cathode 
may tlicrefore be limited either by tem- 

































. — 

- i) 





1 1 

| | 


Degrees /felvip 






Fig. 2. The Effect of Space Charge on the 
Thermionic Currents 

perature or by space charge. With a given 
temperature of the cathode, the thermionic 
current will increase at first as the positive 
potential on the anode is increased, and 
for each voltage V, there will be a Corre- 
ct is evident that equation (3) may be used as a method for 
the determination of elm. The results obtained therefore serve 
to confirm once more the conclusion that the negative current 
from the hot cathode is due to electrons 

sponding value of i s according to equation 
(2). When i s has attained the value i which 
corresponds to saturation thermionic cur- 
rent from the filament at the given tem- 
perature, further increase in voltage has 
no effect. 

On the other hand, with a given voltage 
drop, the current increases with the tempera- 
ture until t is equal to i s , and further increase 
in temperature leads to no corresponding 
increase in thermionic current. This is the 
case illustrated in Fig. 2. 

The existence of this space charge effect is 
evidence of the absence of any positive 
ionization, and serves, therefore, as additional 
confirmation of the conclusion that the 
currents obtained from a hot cathode in a 
very high vacuum are due to a pure electron 
emission, and are not dependent upon the 
presence of any small amounts of gas. 

In this respect the behavior of a hot 
filament in a good vacuum differs radically 
from that exhibited by a Wehnelt cathode. 
In the case of the latter the currents obtained 
are due largely to the presence of positive 
ions, as is shown by the absence of space 
charge effects. The result is that the cathode 
disintegrates under the action of positive 
ion bombardment, and a rectifier containing 
such a cathode therefore cannot be used with 
potentials higher than a few hundred volts 
at most. On the other hand, in the case of a 
rectifier containing a hot filament as cathode 
and exhausted to as high a degree of vacuum 
as possible, there is no conduction except by 
electrons. In order to distinguish the latter 
type of hot cathode rectifier from other forms 
in which positive ions play an essential role, 
the designation, kenotron, has been specially 
coined. This word is derived from the Greek 
adjective kenos, meaning "empty" and the 
suffix tron signifying an instrument or appli- 
ance. The applicability of the name is 

Having indicated the possibility of the 
construction of a high voltage hot cathode 
rectifier, we shall now proceed to discuss the 
principles underlying the designing of such 

Principles of Design of Kenotrons 

The question as to the proper design of a 
kenotron may be treated under three headings : 

1. The amount of current to be rectified. 

2. The maximum permissible voltage loss 

in the rectifier. 

3. The proper form of electrodes to prevent 

electrostatic strains on the filament. 



(1) Current Carrying Capacity of the Kenotron 

The current carrying capacity of a kenotron 
when given sufficiently high voltage between 
the electrodes, is limited only by the area 
of the surface emitting electrons (that is, 
length and diameter of filament) and its 
temperature. The data given in Table I 
and the curve shown in Fig. 1 are therefore 
of fundamental importance in this connection. 
The next consideration is, of course, the 
"life"* of the filament at any temperature, 
and normally the maximum temperature 
at which the filament is maintained should be 
such that the "life" of the filament is over 
1000 hours at least. 

Thus, a 5-mil filament at 2400 deg. K. 
(corresponding to 1 watt per candle) has a 
life of about 4000 hours. The electron emis- 
sion per 1 cm. length of 5-mil filament at this 
temperature, as calculated from Table I is 
15 milli-amperes. The energy required to 
maintain the filament at this temperature is 
about 4.5 watts per cm. length. 

Where the kenotron is required for currents 
of 100 milli-amperes or more, it is better to 
use a 7 or 10-mil filament. This is of advan- 
tage in two respects. Not only is there an 
increase in area per unit length, but also the 
life is much longer at the same temperature. 
On the other hand, the temperature can be 
increased and the life of the filament still be 
maintained at over 1000 hours. Thus, in the 
case of a 10-mil filament at 2500 deg. K., the 
life is pretty nearly 3000 hours, while the 
electron emission is 70 milli-amperes per 
cm. length. 

The data shown in Table II are of great 
interest in this connection. As "safe" 
temperature we consider that at which the 
life of the filament is over 2000 hours. The 
last column also gives the watts per cm. 
length of filament, a figure which is of 
importance in calculating the losses in the 
rectifier itself. 1 


Electron -ut~**- 

Diam. of „..„ 

Filament in T ,„!!i! t „„ Emission per f ^ at T sp ",, t 
Mils Temperature Cm ^^ Cm. Lengtht 



•The "life" of a filament is usually taken in this laboratory 
as the time required to evaporate 10 per cent of the diameter. For 
data on the rate of evaporation of tungsten filaments the reader 
is referred to the paper by Langmuir. Phys. Rev., g, 329 (1913). 

IThese figures are based upon data published by Langmuir, 
Phys. Rev.. Si. 401 (1913). 

tThe current necessary to heat the filame-.t varies from 2 
to 10 amperes, according to the diameter of the filament and the 
temperature, and may be obtained either from a storage battery 
or small transformer. 

(2) Voltage Drop in Kenotron 

Owing to the existence of the space charge 
effect it is evident that for any given current 
carrying capacity i of a kenotron there will 
exist a voltage drop V in the rectifier itself 
and the relation between these will be of the 
form indicated in equation (2). 

We can now consider the manner in which 
the kenotron operates when placed in series 
with a resistance across a source of high 

Let E denote the value of this voltage at 
any instant, and i s the current rectified. 
If V denote the voltage drop through the 
kenotron, and R, the resistance of the load, 
it follows from equation (2) that 

i s= kvi=k (E-i s R)i (2a) 

With constant value of E, the current 
rectified increases as R is decreased until 
i s has attained the value i corresponding to 
saturation thermionic current at the temper- 
ature at which the cathode is maintained. 
If now R is decreased still further, i remains 
constant, and consequently the voltage over 
the kenotron increases beyond that given by 
equation (2). That is, this equation gives the 
minimum voltage drop through the kenotron 
when rectifying a given current i s ; but when 
operating in series with a resistance, the 
voltage drop in the kenotron is that available 
above the i s R drop in load. In case of a short- 
circuit on the latter, where R decreases 
indefinitely, the total voltage of the source is 
taken up by the kenotron, thus liberating 
the whole of the energy, Ei, as heat at the 
anode, and the latter may be raised to a tem- 
perature at which it will melt or volatilize 
and ruin the tube. 

It is necessary to emphasize this character- 
istic behavior of the kenotron, and in practice 
care should be taken to provide against 
short-circuiting of the load, or some form of 
protective device should be used. 

The watts lost in the kenotron owing to the 
space charge effect is 


Yi = kY* 


Because of the high degree of vacuum, 
none of the electrons lose energy by collision 
with gas molecules. The whole of their 
kinetic energy is therefore liberated as heat 
at the anode, just as the energy of rifle 
bullets travelling through a comparatively 
frictionless medium is converted into heat 
at the target. Denoting the number of 
electrons emitted per unit area and per 
unit time by n, and their velocity by v. it 


follows that the energy converted into heat 
at the anode is 

n (J/£ m v 2 ) =n e V =i V (6a) 
If to this be added the watts wh used in 
heating the filament, then the total loss in 
energy becomes 

wl = u>h+wr (7) 

Of this energy loss, the whole of wr and a 
large fraction of wh are used up in heating 
the anode. 

It is evident that if the anode becomes too 
hot the rectification will tend to become 
imperfect. The rectifier must, therefore, be 
so designed that the space charge voltage 
is not great enough to cause heating of the 
anode when the requisite current is being 
carried by the tube. The amount of energy 
(in watts per square centimeter) required to 
maintain tungsten at a temperature T is 
given by the equation,* 

^= i2 - 54 (t4) 47 


Table III gives the values of Ws for 
different temperatures. The last column 
gives the corresponding values of the electron 
emission per unit area in milli-amperes. 



W 5 




1.2 X10" 11 



6 X10-* 










From these data it may be concluded that 
about 10 watts per sq. cm. of anode area is 
quite permissible. This would correspond to a 
temperature of about 1600 deg. K., that is 
a very bright red heat. At this temperature 
the electron emission is still less than 0.02 
milli-ampere per sq. cm. 

(3) Electrode Design 

There remains only one other point to 
consider in the design of kenotrons and 
that is the prevention of electrostatic strains 
on the filament. As is well known, the 
electrostatic force between two charged 
surfaces increases as the square of the 
voltage difference. At voltages of 25,000 and 
over, this force becomes quite appreciable and 
unless special precautions are taken in the 
design of electrodes, it is possible at such 

*I. Langmuir. Phys. Rev.. Si, -101 (1912). The same equa- 
tion is also approximately true for molybdenum. 

voltages to actually pull the heated filament 
over towards the anode. When the kenotron 
is used in series with a load on a high tension 
alternating current circuit, there is a very 
low potential difference between the electrode 
during the half cycle that rectification occurs, 
while during the other half cycle the whole of 
the voltage drop generated by the transformer 
or other source of alternating current occurs 
in the rectifier itself. It is therefore necessary 
to design the kenotron so that the electrostatic 
forces acting on the filament are reduced 
to a minimum. 

Various types of construction have been 
adopted to take care of this difficulty. A 
straight filament in the axis of a cylindrical 
anode; a V- or W-shaped filament placed 
symmetrically between two parallel plates; or 
■ a headlight filament inside a molybdenum 
cap, each of these types of construction has 
been found practicable up to certain voltages. 

Of course, electrostatic forces can be 
overcome by placing the filament at quite 
a distance from the anode and shielding the 
former in the same manner as is done by 
Coolidge in his Rontgen ray tube. But 
under these conditions the "space charge" 
voltage (which increases with the first or 
second power of the distance, see equations 
4 and 5) becomes excessively high and the 
energy loss in such a rectifier would be alto- 
gether too large. 

Different Types of Kenotrons 

The different types of kenotrons mentioned 
in the previous section are illustrated in 
Figs. 3, 4 and 5. In the following section it is 
intended to discuss briefly the characteristics 
of rectifiers that have been constructed along 
these lines and to point out the relative 
advantages and disadvantages of each type. 

Fig. 3 shows a molybdenum cylinder A 
with a coaxial filament F. For direct current 
voltages up to 15,000 the diameter of the 
cylinder need not exceed one-half inch 
(1.27 cm.), while the length may be made 
as much as four inches (10 cm.) A 10-mil 
filament is used as cathode. 

At a temperature of 2550 deg. K. (see 
Table II) the maximum current obtainable 
from such a kenotron is about 400 milli- 
amperes, and the voltage drop necessary to 
produce this current as calculated from 
equation (4), and actually observed, is 


'400 1.27 
' lT6 X 2 X 



= 145. 



The space charge equation for this kenotron 


is = 230 X 10" 3 X V* milli-amperes. 

At 145 volts, \Y R = 145X0.400 = 58 watts. 
Also a/H=72 (Table II). The total energy 
used up in the rectifier is therefore 130 watts. 

energy lost in the kenotron is about 125 watts 
which represents only 1.25 per cent of the 
total energy which the tube is capable of 


Fig 4. Kenotron Containing Cylindrical Anode 

Fig. 3. Molybdenum Cap Type of Kenotron 

As the radiating area of anode surface is 
about SO sq. cm., this energy loss corresponds 
to slightly over 1.5 watts per sq. cm., which 
is just sufficient to maintain the anode at a 
dull red heat (1100 deg. K.). Since the 
kenotron is capable of rectifying0.400 X 15,000 
= 6 kw., the energy loss in the tube corre- 
sponds to about 2 per cent of the total amount 
of energy rectified. 

For direct current voltages up to 75,000 
or 100,000, the diameter of the cylinder is 
increased to about 5 cm. For mechanical 
reasons it has been found necessary, in this 
case, to attach the filament to a molybdenum 
rod framework, which serves to increase the 
space charge voltage above that calculated 
from equation (4). In a tube intended to 
rectify 10 kw. at 100,000 volts the current 
carrying capacity required is 100 milli- 
amperes. This electron emission is easily 
obtained from about 4 cm. of 7-mil filament 
at a temperature around 2400 deg. K. 

The space charge data of Table IV were 
obtained with one kenotron (Xo. 72) of this 

These observations are in accord with the 

*' s =6xio- 3 xn 

energy loss in the tube owing to this 
space charge voltage amounts to 65 watts for 
100 milli-amperes. Adding to this about 50 
watts consumed by the filament, the total 

Fig. 5. Kenotron with Filament Between Two 
Parallel Plates 

A form of kenotron which is suitable for 
voltages not over 10,000 and currents ranging 
up to 100 milli-amperes is that shown in 






33 milli-amperes 

Fig. 4. It consists of a small filament such 
as is used in automobile headlights inserted in 
a molybdenum cap about 1.6 cm. (^ inch) 
in diameter. 

The following table gives the currents 
actually obtained with different voltages in 
the case of kenotrons containing a 7-mil head- 
light filament (No. 50), and 5-mil headlight 
respectively (No. 51). 





























In the case of No. 50, the observations are 
very accurately represented by the equation 

*' S = 34X10- 3 XF 

While in that of No. 51, the corresponding 
equation is 

/ s =i9.5xio- 3 xr* 

In neither case is it necessary to heat the 
filament to a temperature above 2400 deg. 
K. The radiating surface of the anode is 
about 4 sq. cm. and the total energy loss for 
100 milli-amperes is about 50 watts. 

The case of a V-shaped filament between 
two tungsten plates is illustrated in Fig. 5. 

In one case (kenotron No. 66) the plates 
were about 2 cm. apart, while in another 
kenotron No. 70) the plates were twice as 
far apart. Table VI gives the characteristics 
for each kenotron. 

The filament in kenotron No. 70 was about 
7, while that in No. 66 was about 6 cm. 
long.* Each tungsten plate was about 2.5 








































2500 130 





is =24X10" 3 XVI 

is=9.9X10- 3 


♦Owing to lead losses only the central portion of the filament 
was at the temperature indicated. 

X5 cm.; so that the total radiating surface 
was about 25 sq. cm. Kenotron No. 66 could 
be used up to about 40,000 volts, while No. 
70 showed no sparking or straining of filament 
up to 60,000 volts. 

By using a W-shaped 7-mil filament (total 
length about 20 cm.) between two tungsten 
plates 5 cm. square and situated 1.25 cm. 
apart (kenotron No. 54), the space charge 
voltage for given current carrying capacity 
was considerably reduced. The space charge 
equation for this kenotron was found to be 

i 5 =103X10- 3 XV^. 

Owing to the small distance between the 
plates, the filament was not situated exactly 
symmetrically with respect to them, and it 
was therefore not thought advisable to use 
the kenotron with direct current voltages 
higher than 25,000. 

Here again, the energy loss in the kenotron 
for a 10-kw. unit (current carrying capacity 
of 400 milli-amperes) is well below 2 per cent. 

A comparison of the different types of 
kenotrons illustrated above leads to the 
following conclusions : 

(1) For current carrying capacities up to 
500 milli-amperes, either a cylindrical anode 
with a filament down the axis, or a W-shaped 
filament placed between two parallel plates 
may be used. The first named type can 
apparently be made much more efficient' as 
regards losses due to space charge effect. 

(2) Where currents of the order of 100 
milli-amperes or less have to be rectified, and 
the maximum direct current voltage is not 
over 15,000, the molybdenum cap type is 
one that is simpler mechanically and also 
quite efficient. 

(3) For voltages up to 100,000, the. 
cylindrical anode type has proven itself to be 
very practicable and efficient. 



Oscillograms of Performance of Kenotrons with 
Alternating Current Voltages 

In order to illustrate the characteristics of 
a kenotron when used with a-c. sources, 
a number of oscillograms were taken. Film 

maximum voltage ISO. The lower graph 
shows that the rectification obtained was 
absolutely perfect ; also the peaked nature 
of the current wave shows that it was limited 
by space charge throughout the whole cycle. 

Fig. 6. Half-wave Rectification, Upper Curve Gives 

Voltage Over Kenotron; Lower Curve Gives 

Current Rectified. Note the Effect of 

Voltage Limitation 

Fig. 7. Same as Fig. 6. Note the Effect of 
Temperature Limitation 

Fig. 8. Full Rectification, Using Arrangement Shown 

in Fig. 10. Upper Curve — Voltage Over Primary 

of Transformer; Middle Curve — Voltage 

Over Load; Lower Curve — Current 

Rectified. The Latter was Limited 

by Temperature of Cathode 

in Each Case 

Fig. 6 was obtained with kenotron No. 
54 placed directly across the 60-cycle, 122- 
volt terminals. The upper curve represents 
the voltage of the generator, while the lower 
curve gives the current through the kenotron. 
The effective a-c. voltage was 122 and the 

Fig. 9. Same as Fig. 8. The Current Rectified was 

Limited on One Half Cycle by Temperature and 

on the Other Half by Voltage 

It will be remembered that for this kenotron 
the space charge equation as obtained from 
direct current measurements, was 
is = 103X10~ 3 XVK 

It was therefore expected that this relation 
ought to hold quantitatively for simultaneous 
values of voltage and current as measured on 
the oscillogram. The results obtained con- 
firmed this expectation splendidly. 


The following table gives the values of i s 

as observed and calculated for values of V 

corresponding to different intervals of a 
second t after the beginning of the cycle : 





(upper curve) 

(lower curve) 


















A direct current milli-ammeter in series 
with the oscillograph read 68 m.a. 

Film Fig. 7 was obtained with the same 
arrangement of apparatus, but the filament 
temperature was made so low that the maxi- 
mum current obtainable was well below the 
space charge current for 180 volts. The 
current curve begins to flatten at a point for 
which V — 90. The corresponding space charge 
current as calculated from the above equa- 
tion is 88 milli-amperes, while the oscillogram 
indicates 74 milli-amperes. The direct current 
milli-ammeter showed a current of 28 m.a. 

The oscillograms shown in films Figs. 8 
and 9 were obtained with an arrangement of 
apparatus similar to that shown in Fig. 10. 
The low tension side of a potential trans- 
former TT, ratio of coils 20 to 1, was con- 
nected to the 122-volt alternating current 
generator, while the high tension coils were 
connected to two kenotrons AF and A'F' 
as shown in the diagram. (The condenser 
C shown in the diagram was omitted.) 
The direct current was taken from the middle 
point of the transformer and the filaments. 
A load of two 250-volt carbon lamps (60-watt 
type) was connected in series with a milli- 
ammeter and the current strip of the oscillo- 
graph to the terminals BB'. The kenotrons 
used were not of the same construction, 
with the result that the space charge voltages 
for the same current were quite different. 

The upper curve in each film gives the 
voltage over the primary of the transformer, 
the middle curve gives the voltage over BB' , 
while the lower curve gives the current 
through the load. In taking film Fig. 8, the 
temperature of the filaments was maintained 
very low, with the result that both current 
and voltage waves were flattened consider- 
ably. The slight irregularity in the ampli- 
tudes of the two half cycles was due to the 
fact that it was almost impossible to adjust 

the temperatures of the two filaments so that 
they would possess the same electron emission. 
The direct current milli-ammeter read 100 

Film Fig. 9 shows an interesting case in 
which the thermionic current from one keno- 
tron was limited by space charge, while that 
from the other was limited by temperature. 
The d-c. ammeter indicated 140 m.a. When 
taking the oscillogram of the current through 
the load, the voltmeter strip was opened, 
and when photographing the wave of voltage 
over load the current indicating strip of the 
oscillograph was short-circuited. 


Fig. 10. Arrangement for Rectifying Both Half-waves, 
Using Middle Point Connection on Transformer 


Summarizing briefly what has been stated 
regarding the hot cathode rectifier (kenotron) 
it has been shown that : 

(1) The current rectification is due to the 
emission of electrons from a heated filament 
in as good a vacuum as can be obtained. The 
current carrying capacity of the kenotron 
depends only upon the area and temperature 
of the filament, and increases with the latter 
according to an equation of the form : 

-= T 

i = aV Te 


where i denotes the saturation thermionic 

(2) The voltage drop in the kenotron 
depends upon the area, shape and distance 
apart of the electrodes, and increases with 
the current actually rectified according to 
an equation of the form 

i s = k.\" (2) 

where i, denotes the space charge current. 

When * is measured in milli-amperes, the 
magnitude of k varies in ordinary cases from 
5X10 -3 , for very high voltage kenotrons, to 



250 X10~ 3 for lower voltage kenotrons. In 
other words, for a potential drop in the 
kenotron of 100 volts, the rectified current 
varies from 5 to 250 milli-amperes. 

As has been mentioned on page 160. 
equation (2) gives the minimum voltage drop 
over the kenotron when it is operated in 
series with a resistance under most efficient 
conditions. Owing, however, to the fact that 
the filament temperature limits the maximum 
current which the kenotron can rectifv, it is 


B - 

Fig. 11. Arrangement of Four Kenotrons for Making 
Use of Full Voltage of Transformer 

possible for the voltage over the latter to 
exceed the value given by equation (2) as the 
rectifier takes the difference between the max- 
imum voltage available and that consumed in 
the load. Care should therefore be taken in 
using the kenotron to avoid short circuits of 
the load, or some form of protective device 
should be used. 

(3) The actual energy losses in the 
kenotron may be reduced to less than two per 
cent of the "total energy rectified when the 
tube is operated to its full voltage limit. 

(4) Up to the present, kenotrons have 
been constructed for direct current voltages 
as high as 100,000; but there is every expecta- 
tion of being able to extend the field of 
application to 150,000 and even 200,000 volts. 

The maximum current rectified has been 
as much as 1500 milli-amperes (1.5 amperes); 
but it is much more convenient to construct 
these rectifiers in the form of 10-kw. units 
where the voltages required exceed 25,000. 
For lower voltages, smaller units are ad- 

(5) A great advantage possessed by the 
kenotron is that two or more of them can be 
operated in parallel. From the remarks made 
above in connection with equation (2-a), it is 
evident that when a number of kenotrons 
connected in parallel are placed in series 
with a resistance, the current through the 
latter will control the voltage drop and cur- 
rent through each kenotron so that in each 
case an equation of the form (2-a) is satisfied. 

The kenotron thus possesses at least two 
advantages over the mercury arc rectifier; 
firstly, because it may be operated at higher 
voltages, and secondly in the fact that several 
kenotrons can be operated in parallel. 

Applications of the Kenotron 

No doubt a number of applications of this 
device will suggest themselves to electrical 
engineers and physicists. A few words, 
indicating the possible fields of application 
that have already been suggested, will prob- 
ably not be out of place. 

In the physical laboratory where small 
direct currents of a few milli-amperes at 
very high voltages are required, as for spec- 
troscopic work, operating small discharge 
tubes, etc., the kenotron ought to prove 
exceptionally useful. An arrangement similar 
to that shown in Fig. 10, and consisting of two 
kenotrons of the headlight filament type 
with a 60:1 potential transformer will act as 
a satisfactory source of direct current voltages 
up to 4500 or 5000. By inserting a con- 
denser C of sufficiently high capacity between 
the terminals BB' the direct current obtained 
may be made as free from pulsations as 
desired. The kenotron could also be used 
for testing the dielectric strength of insulation 
with high voltage direct currents. 

The writer has obtained as much as 400 
milli-amperes direct current at 6000 to 7000 
volts by using in the same manner a 500-cycle 
generator and a 100 to 10, 000- volt trans- 
former.* By inserting capacity between 
the high tension direct current terminals, 
it was found possible to reduce fluctuations 
in the resulting direct current to less than 
five per cent when 100 milli-amperes was 
being used at 6000 volts. 

Fig. 11 shows an arrangement of four 
kenotrons in which the whole of the voltage 
generated by the transformer is utilized. 
BB' are the direct current leads. 

* The kenotron operates just as satisfactorily on 100.000 
cycles as on ordinary frequencies. 


The combination of kenotrons and trans- 
former could be used to replace the cumber- 
some static machines and the still more 
complicated mechanical rectifiers that are 
at present used to produce high voltage 
direct current for X-ray tubes and the precip- 
itation of dust, smoke, etc. 

Another field of application that appears 
to be very much within the limits of possi- 
bility is that of high voltage direct current 
transmission. While this system has not 
been used to any extent in this country, it is 
a well known fact that the Thurv system has 

met with great success in Europe*. To trans- 
mit 1000 kw. by 100 kenotrons, working in 
parallel at a voltage of 50,000 to 75,000 is 
quite a feasible proposition. 

In conclusion the writer wishes to express 
his indebtedness to Dr. Langmuir and Mr. 
W. C. White of the Research Laboratory for 
valuable suggestions and kind co-operation 
during the work on the development of the 
above device. 

*J. S. Highfield. Journ. Inst. Elec. Eng.. London, SS. 471; 
49, 848; 51. 640. In these papers the advantages of high voltage 
direct current transmission are discussed very fully. 



In preparing the component parts of this article, each author (one representing the generator designer 
and the other the engine designer) has presented his subject with the express purpose of assisting the other 
to a better understanding of the factors that affect parallel operation of a-c. generators driven by internal 
combustion engines, as determined by his end of the set; for it is only through co-operation between the 
builders that generator and engine can be constructed with the correct characteristics to insure satisfactory 
operation when coupled together. Excessive variation in angular velocity, or hunting, is the chief trouble 
to guard against, and the greater part of the article is devoted to a discussion of the natural period, or fre- 
quency, of the units, with the object of avoiding a condition of resonance. — Editor. 


By R. E. Doherty 
Alternating Current Engineering Department, General Electric Company 

The rotor of an alternator which is operat- 
ing in parallel with others will tend to swing 
to and fro at a definite frequency through the 
position of uniform rotation, if the equilibrium 
of driving and resisting forces is disturbed, 
just as a weight suspended by a spring will 
oscillate if the equilibrium of the force of 
gravity and the tension in the spring is 
disturbed. This is an inherent character- 
istic of synchronous machines. Hence, when 
an alternator is driven by a reciprocating 
engine which develops, inherently, a period- 
ically varying turning effort, or driving force, 
there exists as a natural result a possibility 
of unstable operation of the alternator — a 
possibility that the natural frequency at 
which the rotor tends to oscillate may be 
equal to or very near the periodic variations 
of the driving force. This condition of 
resonance, like that of the spring and weight 
receiving impulses in synchronism with 
natural oscillations, will cause swinging, 
or "hunting" of the rotor. The amplitude 
of such rotor oscillations, as measured by 
the maximum displacement of the rotor 
from its stable position (the position of 
uniform rotation) is determined by two 
factors; namely, the magnitude of the varia- 

tion in angular velocity (itself the result of 
periodically varying driving force working 
against the practically constant resisting 
force of load and friction), and the proximity 
to a condition of resonance. A large ampli- 
tude of swing might be produced by a small 
periodic variation in angular velocity, if 
the natural frequency of the alternator is 
very near the frequency of the variation; 
and it is also possible to have a large ampli- 
tude, even if the two frequencies are different, 
if the periodic variation in angular velocity is 
large. Yet, since both of these factors may 
be fixed at predetermined values by the use 
of proper flywheel effect, it is not necessary 
to have either at a dangerous value. 

These facts have been matters of record 
for a long time, having been established in 
the early days of steam engine units. But 
even today a case now and then appears 
where these factors were not properly inves- 
tigated in the design of the unit, especially 
in internal combustion engine units, and 
the usual result in such an instance is excessive 
hunting. In the case of internal combustion 
engines, the additional and more serious 
variations in the turning effort as compared 
with the steam engine unit, make it very 



necessary to consider carefully in each 
instance the natural frequency as well as the 
periodic speed variation, when the flywheel 
is being designed. But of course the engine 
builder, who ordinarily designs and builds the 
flywheels, can not settle with accuracy the 
proper value of natural frequency, unless he 
has at his disposal the generator constants 
on which the natural frequency depends. 

In reviewing the theoretical considerations 
of the problem, and in calling attention to 
some serious operating conditions which 
were found on an investigation of several gas 
engine stations, the object of this article is 
to encourage a further study of the problem in 
general, and bring about co-operation between 
the engine and generator builders in design- 
ing new units. 

In order to develop the relation by which 
the natural frequency may be predicted, 
and to study the limitations of permissible 
variation in angular velocity which are 
required by the alternator, it is necessary 
first to look into the electrical effect of the 
oscillatory movements of the rotor. Suppose 
a generating system is delivering load: an 
alternator is brought up to speed and synchro- 
nized in the ordinary way; and the governor 
or throttle is set so that the wattmeter reads 
zero, that is, the alternator is carrying no 
load. Under this condition the voltage 
generated by the alternator reaches maximum 
and zero at precisely the instants the line 
voltage reaches its maximum and zero values. 
That is, the two voltages are in phase opposi- 
tion at all instants; and if the field excitation 
is adjusted for a value of generated voltage 
equal to the line voltage, then the ammeter, 
as well as the wattmeter, will read zero. If 
the field excitation is adjusted for a higher 
or lower value of voltage, that is, if the alter- 
nator is over or under-excited, the ammeter 
will indicate the resulting wattless current 
required to consume the difference in voltage ; 
but the wattmeter will still read zero. If, 
however, the engine is adjusted so that it 
tends to run at a higher speed, the rotor will 
tend to advance from the position in rotation 
it maintained before adjustment. This of 
course means that the center line of field 
flux has advanced, and that therefore the 
voltage generated by this flux reaches a 
maximum at a relatively earlier instant. 
Hence there has been produced a correspond- 
ing difference in the phase of the line and 
alternator voltages — in the time at which they 
reach their respective maximum values. 
This difference in phase, produced by the 

advance of the rotor, allows current to flow, 
which, by its distorting effect on the magnetic 
field, restrains the advance; and in proportion 
as the displacement from the original position 
increases, the phase difference, and therefore 
the current, also increases, the latter produc- 
ing proportionally increased restraining force 
(the force which tends to restore the rotor to 
the original position). This force, working 
at the peripheral velocity of the machine, 
measures the power input to the alternator; 
and the current produced by the displacement 
is the working current, or the energy current 
of the alternator. Obviously if the displace- 
ment were produced in the opposite direction 
by exerting a drag on the shaft, the alternator 
would be operating as a synchronous motor. 
Hence, to sum up, energy current is produced 
by displacement alone; and this current and 
the force produced by it, tending to restore 
the rotor to the zero position, are both 
proportional to the displacement. Also, if 
the alternator is working at a given load, 
the stable position of the rotor is naturally 
at a given displacement from the zero position ; 
but any tendency to change the rotor from this 
stable position will be resisted, as in the case 
of no load, by a restoring force proportional 
to the displacement from the stable position. 

Fig. 1 

These considerations afford a working 
basis, showing that the motion executed 
by the rotor during an oscillation is harmonic, 
because the restoring force, that is, the 
accelerating force, is proportional to displace- 
ment ; that the natural forces of the alternator 
which characterize it as a synchronous 
machine are the very forces that make it 
oscillate or "hunt" when subjected to per- 


turbing influences; and that aside from the 
natural oscillations, the periodic displacement 
which the prime mover imposes upon the 
rotor, purely by reason of uneven turning 
effort with its resulting variation in angular 
velocity, will produce proportional current 
and power oscillations. 

The relations between voltages and also 
between displacement and restoring force 
are shown diagrammatically in Figs. 1 and 2. 
In Fig. 1, Ei represents line voltage; E„, the 
counter-generated voltage of the alternator; 
0, the displacement angle; e e , the resulting 
cross voltage, which, acting on the impedance 
of the alternator, produces the working 
current ie, almost 90 deg. lagging behind e e , 
and therefore almost in phase with £;. In 
Fig. 2, 6 again represents the displacement 
angle from the zero position 0; and the 
ordinates, /<>, represent the restoring force 
at the different displacements; F r , the force 
corresponding to one electrical radian dis- 
placement, that is, to the displacement 
which would make e e equal to the line voltage. 
(This, for convenience in proportionality, 
carries the assumption that the arc and 
chord of a circle are equal, slightly past 
accurate limits, but the assumption as applied 
involves error only to the extent of the 
difference between arc and chord at the 
angle of hunting, not at one radian. And 
that difference is very small.) 

To relate the factors operating during 
natural oscillations of the rotor, assume that 
any angle 0i is the limit of swing; that is, 2di 

\ ^m\ Fr 

Fig. 2 

is the total amplitude. The restoring force 
corresponding to a displacement 0i electrical 
degrees is 


where F r is the restoring force in pounds 
corresponding to one electrical radian dis- 

The work, W, which will be done on the 
rotor by /» in the movement of the rotor 

during oscillation, through the angle 0, 
is represented by the shaded area, Fig. 2, and 
is equal to 



— X displacement of B\ deg. 


A displacement of 0i electrical degrees rep- 
resents — — feet at a radius of 1 ft., where 
90 q 

q = number of poles. 

Fig. 3 

Substituting in (2), the work becomes 

"-mfc* <3) 

When the rotor swings through the zero 
position, this work will have been trans- 
formed into kinetic energy, 

y 2 MV*- (4) 

where M= mass of rotating element, 

V = maximum velocity of swing in feet per 
second of the center of gyration. Reducing to 
one ft. radius for convenience, M becomes 

]VR 2 

where WR- = weight of. rotating element 

X (radius of gyration) 2 in lb. ft. 2 , g= gravity, 
and V becomes the maximum velocity of 
swing in feet per second at one ft. radius. 

Substituting in (4), the kinetic energy is 

WR*XV*^ ,,_ (5) 


ft. lb. 

But since the motion is harmonic, V may be 
taken as the constant velocity of a point, 
p, moving in a circle whose diameter is 



as shown in Fig. 3. The movement of the 
rotor during oscillation of 20i deg., being 
harmonic, is such that the center line of the 
pole, indicated by the arrow, is at all instants 
under the point p. 

Let T equal time in seconds required by 
the point p to traverse the circumference of 



the reference circle. This of course is also 
the time required by the rotor to make one 
complete swing. Then 

V r =-=rft. per sec. (7) 

T= 0.001705 5 


From (6) and (7) 



45 q T 

ft. per sec. 


Fig. 4 

and the kinetic energy is, from (5) and (8) 

^^^f,lb. (9) 

40o0 g q- T 2 

Equating (3) and (9) 

79 WR- 



Putting g= 32.16 ft. per sec. - and substituting 


where /= generator frequency in cycles per 
second and 

S = R.P.M. 

p = 0.0205 WR* S (12) 

Now, since 

F r = 


33000 P a 


0.746 X 2tt 5 

Fo= Kw. corresponding to the value of e e 
at one electrical radian displacement, 
and the current produced thereby when 
acting across the impedance of the 
machine; that is, the kilowatts corre- 
sponding to normal voltage and short 
circuit current of the alternat' ■-. 

11 'R- S 2 
T* = 0.291 X 10- * „ 7 . and 

\ Pof 

= seconds per oscil- 

lation. Hence the natural frequency in oscil- 
lations per minute is 

F- ^ JM (14) 

V 117?- 

An example will illustrate the application of 
equation (14). 

Fig. 4 shows the saturation curve and short 
circuit characteristic for a 500 kv-a. three- 
phase, 60-cycle, 200-r.p.m. 2300-volt alter- 
nator. At the field excitation, SO amperes, 
which gives 2300 volts on open circuit, the 
corresponding short circuit current is 250 
amperes. These values correspond to 

„ \ 3X2300X250 
Fo= 1000 = 100(Jk ^ 

The combined ITT?' 2 of the flywheel and 
alternator is 2S5.000 lb. ft.-. Hence the 
natural frequency is 

„ 35200 /1000X60 , _ .... 

F ~20TV 285000 = 8 °° osclllatlolls 
per minute. 

The accuracy of equation (14) applied as 
above is probably within 4 or 5 per cent as 
indicated by tests made by a majority of 
investigators, the calculated result usually 
being lower than the actual. The writer 
has had the opportunity of observing accu- 
rately the natural frequency in two instances. 
In these cases the calculated value was 4 
per cent low in one, a 75 kv-a., 60 cycle, 276 
r.p.m. generator; and exactly right in the 
other, a 300-h.p., 60 cycle. 720 r.p.m. synchro- 
nous motor. Whatever error occurs is due 
principally to the fundamental assumption 
that the distortion of the magnetic field 
under load is not affected by the increase in 
reluctance which the distorted field encounters 
in a salient pole alternator.* 

It is of interest to note, in passing, that 
the natural frequency of a given unit is 
independent of speed, and depends only 
upon the magnetic loading of the alternator. 
Because, for a given value of field exciting 
current (which corresponds to a definite mag- 
netic loading when the machine is operating 
on open circuit), the short circuit current is 
practically the same for any speed, except 
zero, of course; and the voltage E a is pro- 
portional to the speed S, as is also the fre- 
quency /. That is, the product, P X/, in 

•For a further studv of natural frequency the reader is referred 
to 'Notes on Flywheel." H. H. Barnes. Jr., A.I.E.E. vol. 23. 
p. 353; "Operation of Alternators," A. E. Everest. J.I.E.E.. 
vol. 50. p. 520; "Parallel Running of Alternators," F. Punga. 
Elek. Zeit. June 11. 1914; "Coupling Flywheel Alternators in 
Parallel." Boucherot. Int. Elec. Congress. 1905. vol. I. p. 692. 


/40 /60 /80 200 220 
f?eYO/ut/ons per- m/nute. 

280 300 

Fig. 5 



equation (14) is proportional to S 2 . Hence 
F must be constant, if the magnetic loading 
is constant, and if this loading is changed — if 
the magnetism in the machine is changed — F 
changes in proportion. For instance, if the 
voltage of a system of alternators in parallel 
is increased, say 10 per cent, by increasing the 
field excitation, the natural frequency of all 
the alternators will be increased by about 10 
per cent. 

Load and power-factor conditions some- 
what modify the value of natural frequency 
as given by equation (14), which, with the 
factor Po as defined, applies to no-load con- 
ditions. The synchronizing-force-per-degree- 
displacement has a different value under no- 
load conditions, as already pointed out. Under 
load and low power- factor conditions this 
force has a different and greater value, and 
the difference is a measure of the change in 
natural frequency. That is, if under load con- 
ditions the synchronizing force is increased 
by, say, 10 per cent, it is equivalent to 
increasing the value of P by the same per- 
centage ; and the natural frequency is changed 
by the extent to which the increased P modi- 
fies the value of equation (14), or about 5 per 
cent. Ordinarily, the change in natural fre- 
quency under load conditions is not serious if 
the voltage is kept reasonably constant. 
Roughly, one can estimate the change by the 
increase in the internal voltage, that is, in the 
magnetic loading. 

Returning to the question of design of 
new units, it is possible to determine P„ 
from the design of the alternator. This 
makes it possible to design the flywheel by- 
equation (14) to produce any desired value 
of natural frequency, and therefore to avoid 
values dangerously near frequencies of the 
engine variations or impulses. Experience 
has shown that if the natural frequency of 
the unit is at least 20 per cent different from 
the frequency of any of the periodic impulses 
of any of the engines in parallel, there will 
be no trouble from resonant hunting. The 
critical frequencies to be avoided are: 

For a four-cycle engine: particularly one- 
half the revolution of the crank, but also the 
revolutions of the crank. 

For a two-cycle engine: particularly the 
revolutions of the crank, but also twice the 
revolutions of the crank. 

It will be noted that the lowest critical 
frequency in either case is the cam shaft 

As an illustration, the danger zones of 
natural frequency for a generator to be driven 

by a twin-tandem, double acting, four-cvcle, 
200-r.p.m. gas engine is 80 to 120, and" 160 
to 240 periods per minute. For a two-cycle 
engine running at the same speed, the danger 
zones would be 160 to 240, and 320 to 480. 
In Part II of this article the causes and 
the relative magnitude of the several engine 
impulse frequencies are discussed. 





\^ c 










Percent offlormol Ltxzcf 

Turning now to the permissible periodic 
displacement of the rotor due to variations 
in angular velocity, it has been shown that, 
regardless of how a displacement is produced, 
it will cause a proportional flow of energy 
current. If a current I , corresponding to 
the power P , flows at a displacement of 
one electrical radian, then for one degree the 
current will be 


Since the permissible value of such a 
pulsating current is properly based on the 
normal rated current of the alternator, the 
permissible number of degrees displacement 
is related to P . For the older type of steam 
engine driven units which were put in service 
before the days of voltage regulators, and 
which therefore had close voltage regulation 
(large value of P as compared with the nor- 
mal rating), the limit was set at ±2.5 electri- 
cal degrees. But for modern units, designed 
for use with regulators, and especially the single 
(maximum) rated generators for use with in- 
ternal combustion engines, P is much smaller 
— of the order of 1.4 to 2.0 times the normal 
rating. For these machines, the permissible 
angle has been increased to ± 3 degrees, which 
would give, in the case of P =l-5 normal 
rating, a pulsating current 

3X . 1 7 5 3 / " = 0.079/„ 

where I„ = normal current. If the generator 
was operating at a power-factor lower than 


unity (and most generators are operated 
under that condition) , the pulsations would be 
somewhat reduced , because there would be very 
little variation in the wattless component. 

Hence the flywheel must fulfill two con- 
ditions: It must limit the periodic variation 
in angular velocity so that the resulting 
displacement will not exceed ± 3 electrical 
degrees, and at the same time must give 
a natural frequency 20 per cent different 
from the frequency of any of the engine 
variations. That is, if it works out that the 
flywheel effect which is required to limit the dis- 
placement to ±3 degrees gives a dangerous 
natural frequency, then the flywheel must be 
increased to remove the natural frequency 
from the danger zone. 

The curves shown in Fig. 5 give the plotted 
results of equation (14). Three values of 
flywheel effect are given for each speed for 
60-cycle alternators, driven by either two or 
four-cycle engines: the upper and lower 
curves give respectively the required value 
of (total WR 2 ■¥ P ) for a natural frequency 
20 per cent below and 20 per cent above the 
lowest engine impulse frequency, which is 
the cam shaft revolutions. The middle 
curve gives the critical value which will 
make the natural frequency equal the cam 
shaft revolutions. 

Yet, the factor of flywheel effect, while 
of primary importance, is not all that must 
be considered if satisfactory parallel operation 
is to be assured. It is a well-known fact 
that if the governors of the several units 
operating in parallel do not have similar 
characteristics (in that the curve of per cent 
speed against per cent load is the same for 
all units), then parallel operation will be 
unsatisfactory because the division of load 
among the several units will not be in pro- 
portion to their capacities. In Fig. 6, curves 
a and b are the load characteristics of units 
A and B respectively. The governor char- 
acteristics determine these curves. For 
this case there is only one point (98 per cent 
speed) at which the units divide load in pro- 
portion to their capacities. At 96.5 per cent 
speed, for instance, A would carry 100 per 
cent, B 125 per cent of rated load. If the 
curves a and b are made to coincide by ad- 
justing the governors, then, of course, a 
proper division of load will occur at all loads 
and speeds. But these points are sometimes 
overlooked. Trouble from this source occurs 
chiefly in stations where engines of different 
manufacture are installed. 

The governor is sometimes the cause of 
trouble of a different sort. When a momen- 

tary change in speed occurs, say at a change 
of load, the governor, by reason of the prin- 
ciple on which it operates, tends to over- 
compensate for the speed change, with the 
result that oscillations will be set up unless 
there is sufficient friction in the governor 
mechanism, or in a suitable dashpot, to 
damp them out. This also is a fact, long 
established, yet now and then overlooked 
on gas engines. 

Nor is this all. It is essential to good 
parallel operation that the adjustments of 
feeding and igniting mechanisms, when once 
made properly, do not change. Poor opera- 
tion would naturally be expected if the 
adjustments were bad to the extent of giving 
an enormous difference in work of the dif- 
ferent cylinders. Experience has shown in 
a number of instances that although the 
parallel operation was satisfactory at the 
time the engine was put in proper adjustment, 
yet after a few weeks of work, large swinging 
of meter needles occurred because the adjust- 
ments had become defective. In many 
instances the operator has no means of 
indicating the engines, and is therefore 
helpless to make accurate readjustments. 
Hence the permanence of adjustments is 
a seriously important point to be considered 
in the design of the engine. 

If all of the above points were considered 
in the design of new units, the parallel opera- 
tion would probably be quite satisfactory, 
without having to take any additional 
precaution in the design of the generator 
over what is taken in the ordinary steam- 
engine driven generator. But, in view of the 
remaining possibilities of the unfavorable 
conditions just described, it is advisable 
for the present to equip the rotor of the 
generators with a low resistance amortisseur 
winding, which dampens any tendency to 
oscillate by consuming as loss the energy 
of the oscillation. However, it should be 
remembered that to get good parallel opera- 
tion by such means, i.e., by overcoming by 
the use of a loss producer on the generator 
the effect of certain features of the engine 
which have not yet been perfected in all cases, 
the object is being accomplished at a constant 
running expense. And therefore in the interest 
of progress and efficiency, as well as of quiet 
meter needles, it seems that by working 
along lines which have been suggested, 
the engine designer could achieve a great 
deal toward carrying further the remarkable 
progress which has already been made in 
perfecting the gas and oil engines for driving 
alternators in parallel. 




By H. C. Lehn 
Snow Steam Pump Company 

In the preceding pages, the electrical func- 
tions involved in the solution of this problem 
are investigated, and their prime importance 
as factors thereof is shown. There is also 
pointed out the desirability of co-operation 
between the generator and engine builder 
in the design of new units, since the generator 
builder only is in possession of the required 
electrical data. It is to be noted further 
that as data of the mechanical factors 
involved are in general available only to the 
engine builder, co-operation is mutually desir- 
able. Generally, the amount of data required to 
be interchanged is small, and its comparison 
will at once determine the most desirable 
flvwheel effect. In some few cases, however, 
the correct solution will not be so readily 
apparent, and more accurate and complete 
data will be desirable, assuming that a gener- 
ating set of the highest possible efficiency 
is to be produced. 

Turning now to a consideration of the 
engine factors of the problem, it has already 
been shown that electrical considerations 
make necessary the limitation of the 
angular displacement by a certain amount. 
The magnitude of the displacement is depend- 
ent upon the variation in the turning effort. 
In particular, its graph is the space curve 
of which the plotted turning effort is the 
corresponding acceleration curve. The latter 
is the resultant of various forces acting in the 
engine, and in its calculation there are met 
a considerable number of factors which do not 
lend themselves readily to the ordinary 
mathematical operations. Hence, the usual 
method of determining the displacement is 
by two graphical integrations of a plotted 
turning effort curve. This procedure requires 
considerable time, and has the further dis- 
advantage of not showing the relative value 
of the various factors, a change in any one 
of them necessitating a complete redrawing 
of the curve. In the present article an analytic 
method will be used, in which these dis- 
advantages are not present, and by means of 
which a comparison of various types of 
engines with regard to displacement and 
natural period will be possible. 

The analytic solution may be arrived at 
as follows: The varying turning effort acting 
on the crank is a periodic function of the time 

and hence, by the Fourier theorem, may be 
expressed in a series of multiple sines and 
cosines. If then there is obtained an equation 
for the acceleration of the rotating parts due 
to the varying turning effort, in a series of 
sines and cosines, the second integral, which 
will be the curve of displacement, can be 
at once written. 

The turning effort equation in this form 
will then be: 

t|^= S (Ai sin qt + A. sin 2qt + 
at iVc 

B\ cosqt-\-etc.) 

from which 

e , A» 

5/= ., ( — A\ sin qt — — sin 2qt.— 
w c q- 4 

B cosqt — etc.) (1) 

in which, 

5/ = displacement in feet on the crank pin 

w c = equivalent weight of the rotating parts 
at the crank pin circle. 

q= where T is the time in seconds of 

the longest forced period (which in the 
present case will be one revolution of 
the engine), so that, 

: 30 ' 

where A T equals rev. per min. of 

the engine. 

g = gravity = 32.2. 

t = time in seconds. 

The variation in speed is very small and 
may be considered constant; hence 0, the 
crank angle, may be put for qt, and since 
only the maximum displacement is required, 
it will be convenient to write Z for the 
maximum value of the series of sines and 
cosines, with their coefficients. 

(1) then becomes 


900 AZ 

W C 7T 2 g W 


where .4= cylinder area in square inches, Z 
being taken on the basis of one square inch. 
Reducing to mechanical degrees : 

, /360\/ 900 AZ\ 

where r = length of crank in feet, and finally 
to electrical degrees for 60-cycle current, and 

5 = 6.03 X 10 s T ;.-^ 
H rN 3 

or in terms of WR-, the flywheel effect, 
5 = 6.03 X 10 8 ,„L l,„ and 

WR? N 3 ' 

WR* = 6.03 X 10 s 


SN 3 



(2) and (3) are in terms of the engine 
dimensions and speed except the factor Z, 
which is determined by the values of the 
coefficients of the terms. These values will be 
initially dependent upon the height and slope 
of the indicator card, and upon the inertia 
force of the reciprocating parts, and will be 
further modified to a considerable extent 
by the number of cylinders and arrangement 
of cranks. If then there is obtained a relation 
between the contour of the indicator card 
and the values of the coefficients, by properly 
combining for each cylinder and for the inertia 
the equation for any type of engine can be 
written. The most convenient form in which 
to establish this relation will be to express 
the coefficients in terms of the indicated 
mean pressures, since the latter is the 
basis of the engine output. With a given 
mean pressure, the slope of the expansion 
line will vary with the clearance volume, 
and with the time of ignition and regularity 
of combustion. The slope of the compression 
line will depend almost entirely upon the 
clearance volume. The latter ranges from 
about 22 per cent of the piston displacement 
for natural gas to about 11 per cent for blast 
furnace gas; and with fairly even combustion 
the exponent of the curve may be taken as 1.3. 
Where there is after-burning, the exponent 
will be lower at the beginning of the stroke, 
increasing more or less regularly with the 
piston travel; and in such cases it will be 
found that the expansion curve agrees closely 
with a curve having a constant exponent of 
1.3 but greater clearance volume, which in 
only a very slow burning card will be greater 
than 25 per cent. Natural gas cards with 

large clearance volume seldom show after 
burning to a great extent, so that a range of 
clearance volume of from 10 to 25 per cent 
should include all fairly normal cards from 
any gas. Such minor irregularities as the 
flattening of the card at the beginning of the 
stroke and the drop of pressure near the end 
do not affect the turning effort appreciably, 
and need not be considered. 

To include oil engines, it would be necessary 
to extend the lower limit of the range for 
clearance volume percentage to about 6. 
Besides, the indicator cards are normally flat 
to an appreciable extent at the beginning 
of the stroke. For these reasons it is best 
to deal with oil engines separately. 

In order to facilitate the forming of the 
equations for any combination of cylinders 
and cranks, it is best to derive the values of 
the coefficients separately for each event of 
the engine cycle. In the derivation of the 
formula, the fundamental period was taken 
as one revolution of the crank and this 
corresponds to the impulse period of a single 
cylinder, single-acting, two-cycle engine, and 
a twin-cylinder, single-acting, four-cycle 
engine, which are the simplest types used 







Per Cent 





Sin 8 
Sin 2 8 
Sin 3 8 
Sin 4 8 

Cos 8 
Cos 2 8 
Cos 3 8 

Cos 4 8 


+ 0.105 
+ 0.006 





+ 0.037 
+ 0.005 

+ 0.437 

+ 0.25S 
+ 0.015 
+ 0.006 

Pt = absolute mean pressure. 
S = crank angle measured from inner dead center. 

for parallel operation. In any single acting 
engine only two events are possible; namely, 
expansion on the overstroke and compression 
on the understroke. In a double-acting engine 
occur the additional events of expansion 
understroke and compression overstroke. It 
develops, however, that the values of the 
coefficients for compression understroke are 
the same as those for expansion overstroke, 
but the cosine terms take the opposite sign, 
and exactly the same relation holds between 
compression overstroke and expansion under- 
stroke. Thus the matter is simplified to 



deriving the coefficients for two events only, 
and for each event for the maximum and 
minimum percentage of clearance volume. 
The values of the coefficients to the first 
four terms for the displacement curve are 
given on the preceding page. It will be 
observed that the relative value decreases 
rapidly with the number of the term, and it 
was found that terms beyond the fourth are 

While it is unnecessary to review in detail 
the principles upon which the construction 
of the turning effort diagram is based (which 
may be found in any text book on mechanics) , 
it may be well to note, that instead of the 

usual form of P for the resultant 

cos <(> 

turning effort of a force P at the wrist pin, 
there has been used its equivalent (within 

sin 2 6) 
negligible error) of P (sin 6-\ ^-. — , where 

the constant I, the connecting rod length 
divided by the crank length, replaces the 
variable <f>, the connecting rod angle. The 
table above is for 1 = 5, the usual value and 
the coefficients will change but slightly for 
other values of /. It is also to be noted that 
the table is based on the absolute mean 
pressures, so that for single acting engines, 
in which one end of the piston is subjected 
to the atmospheric pressure of 14.7 lb. per 
sq. in. (at sea level) there must be subtracted 
from the equation 14.7 sin + 1.47 sin 2 6. 

The equation of the inertia turning effort 
is easily formed. The expression for the 
accelerating force at the wrist pin due to the 
inertia of the reciprocating parts is given in 
all text books, and is: 

K = -0.000341 GrW (cos 6 + c ^ s A e l ( 4 ) 


K = inertia force at the wrist pin per sq. 
in. of piston area. 

G = weight of the reciprocating parts per 
sq. in. of piston area. 

r = crank length in feet. 

AT = r.p.m. 

Multiplying (4) by sin 0+~-7~- gives the 

inertia turning effort at the crank pin per 
square inch of piston area. The resulting 
expression easily reduces to 

Kc = 0.000341 GrN 2 X 
sin 6 sin 2 6 3 sin 
1 2~ 



4Z 2 


for which the corresponding displacement 
equation is 

A' = 0.000341 GrN 2 X 


sind sin2 8 sin 3 6 sin 4 
TT" 1 8 + 'l2l 64 I 2 


By forming from these values of the 
coefficients the equation of the displacement 
curve for the maximum and minimum condi- 
tions, it is possible to determine the variatioh 
of the displacement as the mean pressures 
and contour of the indicator card change, and 
the effect on it of a change in the inertia. 
There can also be noted the effect of a differ- 
ence in mean effective pressure in the separate 
cylinders of a multi-cylinder engine — a condi- 
tion which ordinarily exists, to a small extent 
at least. In determining the maximum dis- 
placement from the equations, the approxi- 
mate crank angle at which it occurs will, 
except in a few cases, be evident from the 
predominating term, and it is then necessary 
to plot a small portion of the curve only in 
the locality of that angle. 

Values of Z, the maximum ordinate of the 
displacement curve for various types of 
engines, are given in Table I. These values 
include a fair, but not excessive, allowance 
for slight differences in the mean pressures 
of multi-cylinder engines, as well as for other 
factors which do not appear in the indicator 
diagram, such as scavenging pumps on 
two-cycle engines, air compressors on oil 
engines, and the varying friction of the 
engine. For this reason the displacement 
figured therefrom will be somewhat greater 
than would be shown by a plotted curve 
where these factors are not taken into 
account. It is found that the difference in 
displacement for the maximum and minimum 
values of cylinder clearance volume is small; 
and for the further reason that indicator 
cards from an engine in operation will show 
some variation in contour as well as mean 
pressure, it has been considered unnecessary 
to include a clearance volume factor in the 
equations. Again, in all types of engines in 
which the predominating torque has a 
period of one revolution, the effect of the 
inertia is very small, since the principal torque 
of the latter has a period of one stroke. For 
the same reason, in such combinations of 
cylinders as result in a predominating torque 


having a period of one stroke, the displace- 
ment is determined largely by the inertia 
forces, and in such cases it is necessary to 
include an inertia factor in the equations. 

The maximum initial displacement has 
been fixed by electrical considerations at 
three electrical degrees either side of mean. 
Accordingly, this value has been substituted 
for 5 in (3), and the resulting constant 
combined with Z is given as C in the table. 

Then WR 2 for three electrical degrees 


from which 

6 = 

N 3 

3 C Ar 


WR 2 N 3 

Both for 60-cycle current. 

It will be of interest to establish a relation 
between the formula for displacement given 
above and that for natural frequency, 
developed in the preceding article, and thus 
correlate the two requirements. For this 
purpose the equation for natural frequency 
(14) in the preceding text will be put in the 
following form: 

273 J 000_ jE^Kl 
N \ WR 2 


in which kv-a. = rating of generator. 

k = short circuit ratio, 
and the constant is for 60-cycle current. This 
formula may be put in terms of the rated 
indicated mean effective pressure P and of 
the engine dimensions as follows : 

I.H.P. E £ l 

Kv-a. = — — — 

1.34 p 


I.H.P. = indicated rated horse power. 
E = mechanical efficiency of engine. 
E l = generator efficiency. 
p — power-factor, 

33,000 ' 
where e is the number of impulses per revo- 

Combining these two equations, substitut- 
ing the value of kv-a. thus found, and also 
substituting for WR 2 equation (3) and 
dividing by N gives : 

Q = 0.075 

P e EE l k 



Q = 

natural frequency 
engine rev. per min. 





Type of Engine Betwe« 


Gas Engines Oil Engines 

Gas Engines Oil Engines 

Single-cylinder ■ o«g 


P \.\P 

2X10 8 P 2.2X10 8 P 


Twin-cylinder ngQ 


1.1P 1.2P 

2.2X10 8 P 2.4X8 8 P 



Three-cylinder 9 .q 


0.7P O.SP 

1.4X10»P 1.6X10 8 P 




Four-cylinder ls( , 


0.5\K-0.23P |0.51A'-0.23P 

1.02X10 8 1.02 X10 8 
(A--0.45P) (K+0A5P) 

Twin-cylinder lgQ 
1 wo-cycle 



(a) 0.24K - 

(b) 0.14P 

4.6X10' 4 - 8 ( %Zl2P) 
(K-0.65P) 2.8XWP 


Single tandem jgg 


(c) 0.14A-- 


2.8 X10 7 

3 X 10 s \/P 

o o 

Twin tandem g Q 

0.056 K+0.09P 

1. 12X10' 
(A- + 1.6P) 




= Rated indicated mean effective pressure. . ..™„v-nnnmii r r \'» 

■ Centrifugal force of reciprocating weights per sq. in. of piston for one crank -0.000341 C r .V. 

With equal pressures in all cylinders. 

For AT. not less than 160 (<:) For K, not less than 1 <o. 

For K. 160 or less (d) For K. 175 or less. 



Values of £, £' and p ordinarily met with in 
practice which would make F a maximum are : 

£ = 0.85; £'=0.92; p = 0.70 which for 

5 = 3 electrical degrees gives 

= 0.139 




and likewise values for a minimum 

■ ■ 

=£cs Engines 

=^W Eng/nes 

Sectioned 'areas /ncf/coCe 
Danger Zor>es 


^ ' '( i I J J l{ 1 y\ 

Lp ! 





i iy> 


; 1 j n r;: r 

r" :' ' j 




! ( I 


5 1 J 


i ..:.. 















I '- 


; | ; 

i 1 I 






\ t $' 

! I ! 





: t ; 

■■; | r 



• \ 





















'. , 













/4 -J<? 






-tC.TCis t RottO' *4£o3.0 for eacfi type 

the natural frequency approaches and passes 
the danger zones, necessitating the considera- 
tion of both frequency and displacement, 
and in the extreme cases, frequency only, in 
calculating the flywheel weight. 

In applying the formulas for displacement 
due regard should be given to the increase 
in the initial displacement as dependent upon 
the ratio of the forced and natural fre- 
quencies, which increase is equal approx- 
imately to ,_^ . , damping neglected, 

*bS -¥o3 X04- ' Yo6 

Fig. 7 

£ = 0.70; £' = 0.S8; p = 0.90. giving 



75 y 



From equations (6) and (7) it will be observed 
that, for any given displacement, the ratios 
of the frequencies are independent of the 
engine dimensions and speed. Equations (6) 
and (7) are plotted in Fig. 7 for the various 
types of engines, and for maximum values of 

P . 

-= in equation (6) and minimum values in (7) ; 

the figures referring to the corresponding 
types in Table I. It will be noted that in 
the simpler forms of engines the natural 
frequency is far removed from the danger 
zones, and that in such cases only the dis- 
placement need be considered, while as the 
number of impulses per revolution increases, 

in which F is the natural and / the 
forced frequency. 

All of the foregoing indicates to 
what extent the solution of the problem 
is a matter of design, while the condi- 
tions occurring in operation which affect 
paralleling are described in the previous 
article. The necessity of proper bal- 
ancing of cylinders, and of designing 
adjusting mechanism so that the adjust- 
ments may be maintained, is referred 
to; and in connection with this it is 
interesting to note that in the case 
of double-acting engines, it has been 
possible in actual cases to reduce the 
displacement by operating with a higher 
mean effective pressure in the under- 
stroke ends. The reason for this fact is 
apparent when the signs of the overstroke, 
understroke, and inertia torques of one rev- 
olution period are compared. 

Again, with regard to governors, it is almost 
entirely a case of adjustment in the field. 
Governor action is hardly susceptible to 
calculation, but analysis seems to show 
the presence of a natural period, which may 
in some cases approach the forced period of 
the engine and thus cause hunting, even when 
the other characteristics of the governor are 
correct. The actual governor period, however, 
is altered by friction, so that accurate pre- 
determination is impossible. 

The general tendency, when not in conflict 
with the other requirements, should of course 
be toward a light flywheel ; for then not only are 
the natural vibrations more quickly damped 
out, but by reason of less bearing friction the 
overall efficiency of the unit is improved. 



By H. E. Spring 
Power and Mining Engineering Department, General Electric Company 

In order to present a convincing argument for the substitution of electric motors for existing steam 
engine equipment on mine hoists it is necessary to be able to show the mine owner some figures on cost of 
operation, as the question of economy is uppermost in any undertaking that is conducted for profit making. 
Complete and accurate tests on motor operated hoists may be obtained with relatively little difficulty, but a 
test of the average steam hoist, to be of value, involves a great amount of work and oftentimes proves to be 
a serious problem. In this article the author recommends a procedure for conducting tests on steam operated 
hoists that will give accurate figures on performance and cost of operation over any desired period of time. 

— Editor. 

There is unquestionably great need of a 
series of carefully conducted tests on the 
various types of large steam hoists operating 
under the different conditions found in 
practice. Conditions obtaining at mines 
are usually unsatisfactory for the economical 
generation and transmission of steam; the 
boiler plants- are usually small and scat- 
tered, and if centralized, steam lines of 1000 
to 4000 feet in length are not uncommon. 
Some properly conducted steam hoist tests 
would most certainly furnish still further 
convincing evidence of the economy and 
other advantages of electrically operated 

A few tests, results of which have been 
published and which really approach actual 
operating conditions, are mostly of foreign 
origin. Tests of too short duration and 
occurring under as near ideal conditions as 
possible are of little value except possibly 
for attractive advertising. Tests true to 
conditions are difficult to carry out in a re- 
liable manner and require prolonged, patient, 
and conscientious effort on the part of all 
concerned. Besides the expense involved, 
the special arrangement necessitated for ac- 
curate results would in many cases hinder the 
work at a busy mine, and this is no doubt 
the reason for the lack of information on the 

The service required of a steam hoist is 
of a distinctive character, and includes among 
other things, variation of speed and horse 
power output from zero to a maximum and 
back to zero again every time a hoist trip is 
made, heavy starting requirements, , and no 
defined length of cycle or interval between 
cycles; these being factors that are not en- 
countered in the tests of steam engines as 
applied to ordinary industrial installations. 

Installations rarely if ever occur where 
only the hoist engine is fed from the boiler 
plant and often the steam line is in common 
with that of other apparatus. It is particu- 

larly essential in testing that arrangements 
be made to secure a self-contained unit 
and at the same time to retain actual operat- 
ing conditions as nearly as possible. The 
conditions under which the rearrangement 
and changes of apparatus around the plant 
must be accomplished are usually not of the 
best. Preparation is more or less handicapped 
because the greater part of the work has to 
be done at night or on Sunday ; and in addition 
to this it is necessary in most cases to keep up 
steam and have the hoist in readiness at all 
times for miscellaneous hoisting or removal 
of the men from the mine in an emergency. 
If a separate man and supply hoist are utilized 
in addition to the main hoist, the situation is 
relieved somewhat and more freedom is 
allowable for rearrangement; but in any case 
due precaution must be exercised at all times 
to avoid any possibility of a serious tie-up 
in the operation of the mine, or the jeopardi- 
zation of human life. 

Because of the pulsating flow of the steam 
taken by the hoist, the measurement of the 
steam consumption by the use of a flow meter 
gives unreliable results. » Even though a flow 
meter were at all applicable it would only be 
effective in measuring the steam for running 
purposes, and would not be sufficiently 
sensitive to include standby losses. On 
account of the clearance space in the cylinders, 
the varying conditions of load, and the vary- 
ing quality of steam, the determination of 
steam consumption from indicator cards is 
only a makeshift method at best. 

The extent to which a steam hoist test may 
be carried is almost unlimited and is deter- 
mined by the anticipated scope and useful- 
ness of the results. The cost of operation is 
usually the main information wanted and this 
may be obtained by taking a short cut, there- 
by sacrificing data which really are not 
necessary in determining the total cost of 
operation, but which would prove of value 
and interest, and might point out possible 



means of improving the existing steam opera- 
tion. Operating conditions at any one mine 
are usually indicative of general practice 
in that particular locality, and an insight into 
methods employed may prove widely useful, 
especially so if electrification is contemplated. 
If it is the intention to determine, in addition 
to the cost of operation, the fuel and steam 
consumption during different periods of the 
day (segregating the idle period as far as 
practical from the active period), the power 
delivered by the engine, the quality of the 
steam, and various other results, very com- 
plete tests and data are necessary. 

No hard and fast rules can be made for 
conducting steam hoist tests because of the 
variance in mine plant practice, different 
types of engines, different kinds of mines, 
and the ultimate results desired. Each case 
demands its own particular solution, and for 
that reason any attempt to set down rules 
covering all tests is impossible. The best 
that can be done is to outline in a very general 
way how the ordinary difficulties can be met 
and the tests carried out. It is hoped that 
the following suggestions, explanations, and 
reasons why some things are done will prove 
of benefit in conducting a steam hoist test 
where complete information is the object, 
and will serve as a guide for any steam hoist 
test whatever the scope of the ultimate 


Proper preparation and forethought will, 
as in any tests, prevent a great deal of con- 
fusion and misunderstanding during the 
tests, and will insure complete data and 
results. The time selected for the test and 
the rearrangements of the plant should be 
such as to permit a continuous .test of at 
least a week's duration at a time when the 
mine is operating under normal conditions. 

The most important and usually the most 
difficult procedure is that of cutting off the 
necessary boiler capacity for operating the 
hoist. The first problem is that of estimating 
the boiler requirements of the hoist, either 
by estimating for the hoist itself, or for the 
other apparatus and leaving the remainder 
for the hoist. Every effort to accurately 
determine and isolate the necessary hoist 
boiler capacity will be well repaid, as the 
temporary changes effected must not ma- 
terially change the ordinary everyday opera- 
tion. In the majority of cases all the boilers 
of the same boiler plant feed into the same 
steam header, and it is possible to blank 

flange sections of this header, thereby seg- 
regating the boilers. The difficulties depend 
on the manner in which the steam lines are 
connected into the header. 

A separate boiler feed pump is necessary 
for the isolated boilers. Any auxiliary ap- 
paratus, such as boiler feed pumps and 
blowers which are necessary for the operation 
of the boilers or hoist, should be fed, if 
possible, from the same system, so that the 
arrangement constitutes a complete self- 
contained steam generating unit and hoist 

Exhaust steam for feed water heaters is 
seldom if ever drawn from the hoist engine 
itself. Other apparatus being the source 
of exhaust steam for feed water heaters, the 
true economy of the hoist engine will not be 
obtained unless the heaters are eliminated 
from the hoist system. Of course there may 
be exceptions to this statement in case the 
exhaust steam can not be utilized for any 
other purpose, and it is a very important 
factor in increasing the hoist engine economy. 
Feed water drawn from the condenser hot 
well of condensing hoist engines must ordi- 
narily receive further heating from heaters, 
and such feed water heaters should receive 
the same consideration as with simple or 
compound non-condensing engines. 

The total steam consumed must neces- 
sarily include all steam chargeable to the rear- 
ranged hoist system, and the only reliable 
way to get accurate results is by measure- 
ment of the feed water. Water meters, as 
a rule, cannot be relied upon for accurate 
work and should only be used as a check on 
other measurements. Means of weighing 
the feed water can easily be provided for by 
the use of two receptacles (tanks or barrels) 
arranged one above the other, the water being 
admitted to the upper receptacle, weighed, and 
then allowed to flow into the lower receptacle, 
to which the feed pump is connected. 

Arrangements for weighing the coal for 
hand firing are easily carried out. One 
means of doing it is by the use of an ordinary 
pair of scales and wheel-barrow. Where 
mechanical stokers are employed, the boilers, 
as a rule, can be hand fired if necessary, and 
the same method pursued as outlined above; 
but more nearty normal operating conditions 
will be maintained if arrangements are made 
to weigh the coal in such a manner that it 
can be fed by the mechanical stokers. Pre- 
caution must be taken in any case to prevent 
use of coal which by accident or otherwise 
has not been weighed. 



The necessary arrangements at the hoist 
engine proper ordinarily make up a small 
part of the total difficulties of preparation. 
The cylinders of practically all modern 
engines have one-half-inch tapped holes for 
making the indicator pipe connections. If 
the holes are not bored, the cylinder heads 
should be removed, if possible, so that the 
exact position of the piston and the size of 
ports and passages may be known; thus 
insuring that the holes will be bored in the 
correct place and facilitating the removal 
of all chips and particles of grit. This 
method involves a great deal of time and 
labor, and probably for that reason would 
not be permissible with a hoist engine. It 
is possible to drill the holes without removing 
the heads by admitting a little steam just be- 
fore the drill penetrates the shell, thus blow- 
ing the chips and grit outward. Care must 
be taken, of course, to protect the workman 
operating the drill. Indicator cards are im- 
portant, but the physical impossibility of 
obtaining them should not interfere with the 
carrying out of the remaining tests. No 
putty or red lead should be used in making 
any of the pipe joints, as particles of these 
materials are liable to cause trouble with the 
indicators. Steam-tight joints can be made, 
if a connection fits loosely, by winding a 
little cotton waste into the threads. If an 
indicator for each end of each cylinder is avail- 
able the piping will need to contain a two-way 
cock for each indicator. Where only one 
indicator is obtainable for each cylinder, a 
three-way cock for each indicator will pro- 
vide the means of transferring from one end 
of the cylinder to the other. 

Up-to-date indicators have a self-contained 
or a separate attachment for reducing the 
motion of that part of the engine from which 
the indicator is primarily driven. The cross- 
head is usually chosen as the most reliable 
and convenient part of the engine for this 
connection. It is hardly worth while explain- 
ing in detail the various accessory appliances 
which have to be made up in the field for 
taking indicator cards, as a little judgment 
and ingenuity will easily . determine the best 
methods for meeting the conditions at hand. 
Single indicator cards are valueless as far 
as the total power developed by the hoist 
engine during a complete hoist trip is con- 
cerned, and therefore continuous indicators 
must be utilized. A detailed description of 
the parts and the operation of continuous 
indicators is unnecessary, as such information 
is always accessible in engineering handbooks 

or can be obtained by application to the 

The usefulness of continuous indicator 
cards depends on the record of the engine 
speed in r.p.m. kept at regular intervals 
during the time the cards are taken. The 




CONTACT /"JflK£»< 









Fig. 1. A Convenient Arrangement of Apparatus for Taking 
Continuous Indicator Engine Diagrams 

arrangement shown in Fig. 1 will prove 
convenient and reliable for recording the 
speed graphically. The spring motor feeds 
the paper along at a uniform rate; A and B 
are pens actuated by magnets; A is con- 
trolled by the time clock and divisions repre- 
senting time in seconds are recorded on the 
paper; B is operated by the contact-maker 
and every stroke of the engine is indicated 
on the paper. 

The contact maker is shown operated by 
the tail rod of the engine, but it can be placed 
at any other convenient place; the more 
contacts made per revolution the better, 
especially with a low speed engine. From 
the complete record on the paper, speed- 
time curves can be plotted. The signal bell 
can be displaced by some other means 
of signalling, if advisable, as the amount of 
signalling will not be great. The stroke 
counter is of use in recording the total number 
of engine strokes and serves as a check on the 
graphic instrument; it is also useful in estab- 
lishing and checking the number of hoist 
trips for various periods, particularly when 
the hoisting is all done from one level. 

Steam pressure records, temperature read- 
ings, and determination of quality of steam in 
the hoist house will require various tappings 



into the steam line for gauge connections, 
thermometer wells, etc.. and such apparatus 
should be located as near as possible to 
the hoist engine. The same general prepar- 
ation applies for obtaining boiler plant 
pressure and temperature readings of the 
feed water and steam. Readings of tem- 
peratures, pressures and quality of steam are 
not of extreme importance as far as the hoist 
test results are concerned, but will prove 
interesting information and could possibly 
be used in getting at approximate operating 
characteristics of the boiler plant alone. A 
little judgment of the conditions will decide 
whether the information is worth the time 
and labor required. 

The application, limitations, and operation 
of calorimeters for determining the quality 
of steam are carefully explained in the various 
engineering handbooks and in manufacturers' 
catalogues or pamphlets, and therefore will 
not be dealt with here. It is assumed that a 
throttling calorimeter can be used; but if 
the percentage of moisture is very irregular 
and is in excess of three per cent, a separator 
must be used in connection with the throttling 
calorimeter, or else a separating calorimeter 
substituted for the throttling calorimeter and 

It is essential that the records be complete, 
that the readings be consistent, and that 
they bear definite relation to each other. 
One great asset in promoting this is by prop- 
erly prearranged log sheets. The required 
number can easily be prepared before the 
tests, by means of carbon copies or by 
mimeograph. Plenty of blank space should 
be left for possible changes, additional data 
and remarks. A sufficient number should 
be provided so that each day's tests can be 
put together and kept separate from the 
succeeding day's tests, as this will be advan- 
tageous in working up the tests. 

The following headings for log sheets will 
serve as a guide, and can be modified to fit 
the conditions encountered: 

Coal, Steam Pressure and Temperature 

1. Time. 

2. Coal consumed. 

Number barrow loads. 
(b) Weight in pounds. 
a pressure. 
4. Steam temperature. 

Feed Water 
1. Time. 

umber receptacles measured. 

3. Weight water in pounds. 

4. Feed water temperature. 

5. Remarks. 







Number trips and origin. 






Men (No. trips, weights and origin). 


Material (No. trips, weights and 


Waste (No. trips, weights and origin). 












3. Indicator cards. 

(a) Number of card. 

(6) Weight hoisted and classification. 

(c) Stroke counter readings. 

4. Remarks. 




(a) Beginning of trip. 

(6) Time for acceleration. 

(c) Time for retarding. 

(d) End of trip. 

(e) Rest period. 


Maximum engine r.p.m. 


Total revolutions of engine. 


Load and classification. 




1. Time. 

2. Gauge pressure in steam line (p). 

3. Gauge pressure in calorimeter (/>,)■ 

4. Atmospheric pressure (/>„). 

5. Temperature in calorimeter (/,-). 

6. Absolute pressure in steam line (P). 

7. Absolute pressure in calorimeter (P.-). 
S. Total heat corresponding to P, (X). 

9. Heat of vaporization corresponding to P (r). 

10. Heat of liquid corresponding to P (Xq). 

11. Temperature of saturated steam correspond- 

ing to P . 

The above headings can be rearranged or 
combined to give the most convenient ar- 
rangement in taking the data. 

* Priming = 1 — 

•v--»s < -t„)-a 



By productive hoisting is meant the hoist- 
ing of coal, ore, or whatever material the 
mine derives its revenue from. Non-pro- 
ductive hoisting includes all remaining hoist- 
ing, such as men, material, supplies and waste. 
Origin refers to the point from which the 
hoist load comes, and applies mainly to a 
multi-level shaft or slope mine. If two 
recording pressure gauges are available, 
readings from indicating gauges can be 
eliminated except for calorimeter tests. 
Graphic records of pressure at the boilers 
and hoist engine give a much better picture 
of all-day operation. 


All testing apparatus, steam and water 
lines, and other rearrangements should be 
thoroughly inspected in order to make sure 
that everything is in shape for accurate 
results. Apparatus for weighing or measuring, 
such as scales, tanks, gauges, etc., must have 
their accuracy established before beginning 
operations, as well as occasionally during 
the tests. 

Preliminary sample indicator cards should 
be taken, as they will show the operating 
characteristics of the engine and may be the 
source of permanent improvement of the 
engine economy through resetting the valves. 
Such discrepancy in valve setting is rather 
remote, but if it exists and it is deemed 
advisable by the owner to remedy it, much 
time and energy will be saved by its detection 
before starting a series of tests. 

The time of beginning the test is not very 
important, since the total time should con- 
sume at least a week, thereby taking account 
of day-to-day conditions. In any case the 
test should include Sunday, as a better con- 
ception of standby losses is then obtained 
than at any other time. 

In order to associate one reading with 
another, readings for the different sections 
of the rearranged plant must all be taken at 
the same time, and the simplest method is to 
read on the hour and half-hour. Consistent 
and associated readings provide means of 
segregating the various periods from each 
other, make ' the tests complete for this 
period, and thus permit of definite results 
being obtained for any particular period 

The depth of the fire in the boilers can be 
estimated at the beginning and the same 
depth approximated as near as possible at 
the end. An assistant will have to be in 
constant attendance to oversee and register 

the amount of coal consumed. If wheel- 
barrows and a pair of scales are used, an 
average wheelbarrow load can be weighed 
and the scales locked at this weight; succeed- 
ing loads are then added to or reduced to 
meet this predetermined weight. Approxi- 
mately the same amount of weighed coal 
should be maintained before the boilers; the 
quantity depends on conditions, but should 
be definitely known at the beginning and end 
of the test. In order to avoid error, every 
time a load is weighed a notation should be 
made; the total weight per half hour being 
computed from these notations. 

The height of the water in the boilers as 
shown by gauge should also be noted and 
this water level maintained throughout the 
tests, so that the weights of feed water will 
line up consistently with the other data for 
any period. 

The duties of weighing and recording the 
amount of feed water require the attention 
of two assistants, one for weighing the water, 
and the other for feeding the water properly 
to the boiler. The receptacle, whatever it 
may be, to which the feed water pump is 
connected should be kept filled to approxi- 
mately the same level all the time. The 
receptacle in which the feed water is weighed 
must not be of too great a capacity com- 
paratively, as the weights may appear in- 
consistent when segregating a short period 
of operation. 

Continuous indicator cards represent the 
total work done by the engine, and cards 
should be taken for each distinctive condition 
of load and speed under which the hoist 
engine operates. Diagrams for special con- 
ditions, such as for determining the frictional 
losses of the hoist, head sheaves, and shaft 
itself, will also prove instructive and valuable, 
and opportunity for obtaining such cards 
should not be neglected. It is necessary 
that diagrams be taken on both ends of each 
cylinder, as a satisfactory card from one end 
does not prove in any way that like con- 
ditions prevail at the other end. A contin- 
uous indicator for each end of each cylinder, 
when operated simultaneously, gives the 
whole story. 

Referring to Fig. 1, the method of taking 
cards simultaneously is as follows: The 
spring motor is started and the man in charge 
signals his assistants to get ready for the 
next hoist cycle; he then starts the time clock, 
and also makes sure that the circuit for B 
is all o.k. 'All indicators can be put into 
working order with the exception of the con- 



tinuous feed before the above signal is given. 
At the signal which occurs before the hoist 
starts, the continuous feeds are turned on. 
Succeeding cycles may now be taken without 
further adjustment, unless the time between 
cycles makes it undesirable to let the clock 
and spring motor run continuously. Record 
the reading in log sheet of the stroke counter 
at the beginning and end of each cycle, as 
this serves as a check on the record made by 
B. Also note time at which cards are taken, 
the weight of productive, or non-productive 
material hoisted or remarks concerning con- 
ditions, and the number of each card. A 
simple system of numbering on the time 
records and indicator diagrams themselves 
will establish their relation to each other. 
Notation must also be made on the indicator 
diagrams to show whether they were taken 
on head end or crank end, and whether left 
or right cylinder. One assistant can manip- 
ulate both indicators on one cylinder, as the 
difficulties of starting and stopping the 
indicators simultaneously, such as experienced 
with constant speed engines, are eliminated; 
no record being made on the cards until the 
engine starts, and the record ceasing when 
the engine stops. 

In case only one indicator per cylinder is 
available, trial continuous cards must be 
taken simultaneously on both ends of the 
cylinder. From these cards a definite ratio 
of power delivered by head and crank ends 
of the same cylinder is established, and in 
order to accomplish this result two indicators 
will have to be temporarily attached to one 
cylinder. This ratio can be determined 
approximately also by taking a continuous 
card on one end of the cylinder during one 
cycle, and then with the same indicator 
taking a continuous card on the other end of 
the cylinder, during a cycle when the condi- 
tions are practically the same as those at 
the time the first continuous card was taken. 

A sufficient number of cards should be 
taken each day to be certain that day-to-day 
conditions are covered, as well as to make 
sure that the operating characteristics of the 
engine are maintained constant. 

Unless extremely difficult conditions pre- 
vail, one assistant can take the readings 
classified under "Hoist Duty." The magni- 
tude of the daily record of output kept 
at the mines by the mining companies varies 
greatly. The quantity of data and the 
difficulties of getting authentic data depends 
entirely on whether the mine is opened by a 
shaft or by a slope; whether hoisting is all 

done from one level or from several levels; 
and if a slope, whether a skip or several cars 
are used for transporting the product. If a 
shaft, the number of levels will not be great; 
they will be definitely located, because of 
their distance apart, and a complete record 
will be comparatively easy to obtain. If a 
slope, it will be exceptional if hoisting is all 
done from one loading station; usually nu- 
merous levels or headings situated at rather 
irregular intervals branch off on either side 
of the main slope, each in itself constituting 
a loading station; consequently a full under- 
standing of where each trip comes from 
requires a rather elaborate record. The total 
number of the productive trips per day, and 
the total weights of product with its segrega- 
tion into quantity per level, if more than one 
level, will ordinarily be obtainable from the 
tipple record. The total productive trips 
for each period can also be taken from the 
tipple record, if it is consulted every half hour, 
and by a little pressure brought to bear on the 
right place it may be possible to obtain the 
origin of the product as well; otherwise the 
origin must be obtained by observation, as will 
also the information concerning non-produc- 
tive hoisting. The non-productive weights will 
have to be estimated. The total daily mine 
record will serve as a check on the total daily 
readings obtained by observation. 

The taking of indicator cards, sampling 
coal, and steam calorimeter tests can be filled 
in between the regular half -hour readings, 
thus making it possible to obtain some help 
from the assistants taking the half-hour 
readings. At night, or during any other 
period outside of the regular hoisting period, 
it will be possible to double up on the keep- 
ing of records and arrangements made 
whereby the force can be materially reduced, 
probably only two men being required. 

"Miscellaneous Hoisting Observations " 
give a very good idea of how the hoist is 
operated during light and heavy hoisting 
periods, with regard to maximum rope speed, 
total running period and rest period. These 
observations may be dispensed with in case 
the indicator cards and time records are 
sufficiently complete to supply the informa- 

Before recording any readings for deter- 
mining the quality of steam, live steam should 
be admitted to the calorimeter for at least 
ten minutes, to insure the temperature of 
the instrument coming to full heat. When 
all is ready, take the following readings 
simultaneously: p, t c , pc and p a \ P and P c 



come directly from readings taken; /„, A', 
r, q and t a are taken from steam tables. 
Sufficient sets of these readings should be 
obtained at various times and under various 
conditions to cover any contingency which 
might arise. 

Sampling coal for moisture and heating 
value is a rather extensive process, and if 
carried out, a handbook or some other source 
of detailed information should be consulted. 
Samples for moisture can be taken every day, 
while two samples for heating value and 
analysis, tested in two separate laboratories, 
will suffice. 

There is a certain amount of miscellaneous 
information which is necessary in getting 
at the total cost of operation (which is usually 
the ultimate object in view) ; also for getting 
at steam consumption per unit of useful 
work and various other results. As regards 
the boiler plant, this data should cover the 
entire boiler plant from which the boilers 
for operating the hoist were segregated. It 
will usually be easier to obtain the data in 
this way, and from this determine what 
belongs to the hoist account. Practically 
the only way of arriving at the boiler plant 
operating costs chargeable to the hoist is by 
applying the cost per ton of burning the coal 
under regular conditions to the coal burned 
during the test. It may not be possible to 
obtain installation costs, labor costs, coal 
burned, and maintenance, repairs and supplies 
as outlined below, because of the various ways 
of accounting, and in that case it is necessary 
to make the best of what is available. Some 
of the data may appear superfluous, but it is 
better to have too much information than not 
enough. The following notes, as well as any 
useful pencil diagrams or layouts, should 
therefore be taken at some time during the 

Boiler Plant 

Installation cost and present value. 

Type and make of boilers. 

Rating in horse power and dimensions. 

Grate surface. 

When installed and present condition. 

Mechanical stokers used (if any). 

Economizers or feed water heaters (if any). 

Feed pumps (type and size). 


Boiler house, cost, dimensions, etc. 

Boilers blanked off 

Accessories blanked off. 

Diameter, length and condition of steam main 

to hoist. 
Kind and price (delivered) of coal burned. 
Number of firemen and wage. 
Number of ash handlers and wage. 

Number repair men and wage. 

Maintenance, repairs and supplies for past six 

months or year. 
Tons coal burned during past six months or year. 

Engine and Hoist 

Installation cost and present value. 
When installed and present condition. 
Type and make of engine. 
Size of steam cylinders. 
Diameter of piston rod. 
Kind of valves. 
First, second or third motion. 
Type of drum (cylindrical or conical, etc.). 
Single or double drum; clutched or fixed. 
Drum dimensions. 
Balanced or unbalanced hoisting. 
Type of brakes and how operated. 
Dimensions and value of building. 
Number of hoist engineers and their wage. 
Engine and hoist maintenance, repairs and 
supplies for past six months or year. 


Total length of haul. 
Length of haul (ground level). 
Inclination to horizontal. 
Number of levels. 

Location of levels (profile of slope or shaft). 
Weight of cage. 
Weight of car. 
Weight of skip. 
Number of cars per trip. 
Size of rope. 

Condition of shaft or slope. 
Tons productive for past six months or year. 
Tons non-productive for past six months or year 

Calculations from Tests 

It is first necessary to arrange and con- 
dense the data into the most convenient form 
for quickly arriving at the final results and 
for making up the report. Time will be saved 
by arranging the data so that the day and 
night, or hoisting and idle periods, can be 
totaled separately for each 24 hours. The 
idle period referred to includes all time 
besides what is considered the regular hoist- 
ing period, or periods. Tables for each 
24-hour results, with the following headings, 
are suggested. 

Boiler Plant and Steam Readings 





(a) Pounds consumed. 

(b) Heating value. 

(c) Moisture. 


Pounds feed water. 


Steam pressure. 

(a) At boiler plant. 

(b) At hoist engine. 


Steam temperature. 

(a) At boiler plant. 

(b) At hoist engine. 


Quality of steam. 





Hoist Di 1 v 





(a) Productive. 

(ft) Non-productive. 



(c) Total. 


Weight in tons. 

(a) Productive. 

(ft) Non-productive. 



(c) Total. 


Average haul in feet. 

(a) Productive. 

(ft) Non-productive. 


Average vertical lift in feet. 

(a) Productive. 

(6) Non-productive. 


Horse power hours net work. 

(a) Productive. 

(ft) Non-productive. 

(c) Total. 

i . 


Indicator Cards 




Card number. 


Weight hoisted, or lowered, and classification 


Length of haul in feet. 


M. E. P. 

(a) H. E. each cylinder. 

(ft) C. E. each cylinder. 


Total revolutions of engine. 

1 . 

Average r.p.m. 


I. H. P. hours. 

a 1 H. E. each cylinder. 

(ft) C. E. each cylinder. 

(c) Total for engine. 



Every item under "Boiler Plant and Steam 
Readings" is available directly from the log 
sheets, with the exception of heating value of 
coal, moisture in coal, and priming of the 

The results covered by "Hoist Duty" re- 
quire rather extensive calculations, especially 
if hoisting is not all done from one level. 
The "Average Haul," when hoisting from 
several levels, is determined for the period 
by dividing the sum of the net weights hoisted 
multiplied by the distance hauled, by the 
total net weights hoisted for the period. 
If a vertical shaft, the "Average Vertical 
Lift" is the same as the "Average Haul." 
Horse power hours net work = 

Net weight in lb. hoisted X vertical lift 

The continuous indicator diagrams should 
first have lines drawn perpendicular to the 
-pheric line through the points which 
mark the ends of the strokes. If an ordinary 
planimeter is used, the area of each individual 
diagram must be taken separately and 

then all added together, taking cognizance of 
positive and negative areas; but, if an 
integrator is put into service, the resultant 
area of the continuous diagram can be 
obtained direct from the planimeter. The 
mean effective pressure in either case is 
obtained by dividing the resultant area by 
the product of the length of an individual 
diagram (shown by the vertical lines) and 
the total revolutions, and then multiplying 
this average height by the scale of the 
indicator spring. Now by multiplying 
together the mean effective pressure, effec- 
tive area of piston in square inches, length 
of stroke in feet, and total number of revo- 
lutions, the result will be foot-pounds of 
work produced by one end of one cylinder 
during a hoist cycle; and this result divided 
by 60X33,000 gives indicated horse power 
hours. The sum of the indicated horse 
power hours of the different cylinders is 
the total indicated horse power hours per 
trip. The total number of revolutions per trip 
is taken from the graphic speed record and 
checked by the stroke counter. Providing 
that a sufficient number of cards have been 
taken, it will be possible to determine closely 
the indicated horse power hours required for 
each kind of trip made, and from this the 
approximate total indicated horse power 
hours per day, or for any period of the day. 

The individual diagrams on the continuous 
cards may show reverse power, and if so 
indicate that the engine is being "plugged" 
during retard for the purpose of bringing the 
hoist to rest without applying the brakes. 
So-called "plugging" is brought about by 
reversing the engine and admitting steam. 
The piston, however, is going in the reverse 
direction to that of the steam, and the result 
is that the steam and air confined in the 
cylinder are finally forced back into the steam 
line. Usually the overall economy is not 
affected materially by this operation, be- 
cause no steam is exhausted to the atmos- 
phere, and the only detrimental effects accrue 
from radiation and the admission of cold air 
through the exhaust ports and thence into 
the steam line. Another method of retarding, 
which is sometimes used but which is not as 
effective as "plugging," consists of throwing 
the valve gear on the central position (point 
of no valve movement) thereby closing all 
the ports and causing the air and steam con- 
fined to be compressed and expanded alter- 
nately. The only retarding effect resulting 
from such practice is that incident to the 
difference of power required to compress the 



Fig. 2. Continuous Indicator Diagram of Twin Simple Hoist Engine 

air and the power obtained from it by ex- 

Fig. 2 shows a continuous indicator card 
taken on a twin simple hoist engine. 

Fig. 3 gives the curves that were made up 
from the set of cards, including the one shown 
in Fig. 2. Typical curve sheets such as these 
should be made up in addition to the above 

The final test results can be expressed in 
several different ways, depending on what 
unit basis is used and on the detailed results 
desired. Some of the following items sug- 
gested for making up the table of final results 
may not be desired, but nevertheless the list 
will give the form of the various results ob- 
tainable. The totals for the complete test 
should be given at the bottom of this table. 





ie ib so sa e* 

Fig. 3. Curves Plotted from Continuous Indicator Diagrams 


Time in hours. 

(a) Hoisting. 

(6) Idle. 
Tons hoisted. 

(a) Productive. 

(6) Non-productive. 

(c) Total. 
Pounds coal consumed. 

(a) Hoisting time. 

(6) Idle time. 

(c) Total. 
Pounds feed water. 

(a) Hoisting time. 

(b) Idle time. 

(c) Total. 

Indicated horse power hours. 

(a) Productive. 

(6) Non-productive. 

(c) Total. 
Horse power hours net work. 

(a) Productive. 

(b) Non-productive. 

(c) Total. 

The test records will also permit still 
further segregation of the active hoisting 
period or the idle period so that results con- 
cerning some distinctive period of the day 
can be obtained. 

The Report 

The value of a report does not depend on 
its length, but on convenient arrangement, on 
the brief, concise statement of facts and 
results, and on the curves and diagrams in- 
cluded. The very first part of the report, 
after a general idea of the conditions has been 
obtained, should be the summary of results, 
the detailed information making up the latter 
part of the report. The best arrangement is 
subject to personal opinion, but the following 
outline with accompanying explanation is a 
very good arrangement: 

General. — Brief introduction concerning 
location of mine, product mined, and normal 
method of operation. 

Object of Tests. — A short, straight-to-the- 
point statement of what ultimate result, or 
results, the tests were made for. 

Existing Conditions. — List of apparatus 
with its condition and arrangement before 
rearrangement for test was effected. The 



data given under "Procedure" for "Mine" 
should also be included here. 

Rearrangement for Tests. — List of rear- 
ranged apparatus, specifying its arrangement 
in conjunction with the plant itself and with 
the extra test apparatus. 

Procedure. — A brief outline concerning the 
method of running the test. 

Summary and Conclusions. — The length and 
scope of the summary depends on how much 
detail is desired. The following list of results 
for the complete test may be revised to suit 
conditions and requirements. 

Total hoisting time in hours. 
Total idle time in hours. 

Total length of test in hours. 
Total number productive trips. 
Total number non-productive trips. 

Total number trips. 
Total productive tons hoisted. 
Total non-productive tons hoisted. 

Total tons hoisted. 
Total pounds coal consumed during hoisting time. 
Total pounds coal consumed during idle time. 

Total pounds coal consumed. 
Total pounds feed water during hoisting time. 
Total pounds feed water during idle time. 

Total pounds feed water. 
Total indicated horse power hours productive. 
Total indicated horse power hours non-productive. 

Total indicated horse power hours. 
Total horse power hours net work productive. 
Total horse power hours net work non-productive. 

Total horse power hours net work. 

Pounds coal per hour hoisting time. 

Pounds coal per hour idle time. 

Pounds coal per ton hoisted. 

Pounds steam per ton hoisted. 

Pounds steam per pound of coal consumed. 

Pounds steam per indicated horse power hour. 

Pounds steam per net horse power hour. 

Heating value of coal. 

Moisture in coal. 

General quality of steam. 

The conclusions drawn depend on con- 
ditions met with, but are ordinarily con- 
fined to ways and means of possible increase 
in economy in operating the plant and to an 
explanation of the test results, which are not 

Detailed Report. — This part of the report 
is simply for reference in getting details con- 
cerning the results given in the summary. 
All the tables made up in the form given under 
the section "Calculations from Tests," should 
be included here. Sample copies of log sheets 
and indicator cards may also be included if 
advisable. Tables of capitalization and 
operating cost should not be incorporated 
in the report unless the final results and 
summary cover cost of operation. 

Curves. — Almost innumerable curves can 
be made up. The most instructive will 
probably be a curve sheet for each day's 
operation plotting pounds coal consumed, 
pounds feed water used and horse power hours 

52°PM 5X*tm?2!PM 
Sun Mon *tan. 

Fig. 4. Curves showing Engine Performance as Determined by One Week's Test 



net work, against time. A curve sheet such 
as this can also be made up for the total test 
using each day's results as a point. Fig. 4 is 
illustrative of curves made covering a week's 

No doubt the lists of test data, etc., have 
appeared to be of a more detailed nature than 
necessary, and have included data which 
have nothing to do with the hoist test itself. 
Such data and information are mentioned, 

however, because of their value and applica- 
tion to the final test results. With this end in 
view, the final results can now be used, 
in conjunction with the data at hand, for 
obtaining the total cost of operation per year 
for this particular mine; and all the results 
thus possible to obtain can in turn be applied 
to other mines, providing proper cognizance 
is taken of the idle time, tonnages, and the 
various other vital points involved. 


By E. P. Peck 
Asst. Electrical Engineer, Georgia Railway and Power Company 

The ordinary telephone instrument, with its fine wire coils, contacts, etc., is a very delicate instrument 
and when used on lines paralleling high tension transmission lines requires a protective device that will effec- 
tively shield it from the abnormal stresses resulting from a cross between the telephone line and the trans- 
mission line, a stroke of lightning, etc. This article describes a telephone lightning arrester, built in three 
sizes, which will adequately protect the instrument on transmission systems operating at voltages up to 
250,000, or higher.— Editor. 

The protection of telephones and other 
terminal apparatus connected to telephone 
lines paralleling high voltage power lines 
has been a very serious problem. The require- 
ment is that the delicate telephone windings 
of approximately 0.005 wire, hook switch 
contacts, etc., with very close spacings, 
must be so protected that they will remain 
in good operating condition after an almost 
unlimited voltage has been repeatedly applied 
to the lines to which the telephone is con- 

This extreme requirement has apparently 
been fulfilled by an arrester that has been 
designed and that has stood tests and operating 
service which seem to prove that it will give 
the telephone good protection when voltages 
of any value or frequency are applied to the 
telephone lines. So far we have found but 
one exception: When the power impressed 
on the telephone line is not sufficient to blow 
a five-ampere fuse, but with voltage high 
enough to keep the arrester continually 
discharging, the telephone equipment will be 
eventually damaged. An explanation of this 
will be given later in the article. 

The telephone high voltage arrester, which 
is designed for voltages from 33,000 to 250,000 
or higher, is satisfactory, as far as protection 
is concerned, for use on any telephone, but 
its size and cost prohibit its use on lower 
voltage lines. For this reason a smaller 
arrester is being designed for use on telephone 
lines paralleling power lines of from 2600 
to 35,000 volts, and another one has been 
made for use on lines which are not subjected 
to higher voltages than 2500. All of these 
arresters apparently give thorough protection 
from any instantaneous application of high 
voltage, such as a stroke of lightning. The 
high voltage arrester was mentioned by 
Mr. C. E. Bennett, Electrical Engineer for 
the Northern Contracting Company, in an 
article in the December number of the 
General Electric Review. 

About three years ago it was necessary to 
protect some telephone lines which were 
subjected to crosses with a 22,000-volt 
power line. An arrester, shown in Fig. 1, 
was made up of an old marble slab, glass 
tube expulsion fuses, and some other material 
which was on hand. It was our intention to 



try out the practicability of this arrester 
with as small a cost as possible and later 
build one with better mechanical arrange- 
ment. In this arrester the line wires con- 
nected at the top and the telephone wires 

Fig. 1. First Experimental Arrester 

and ground wires at the bottom. A spark 
gap of 0.004 inch between knurled brass 
cylinders was placed from line to line on 
the telephone side of the fuses. This gap 
was set very close for the reason that it was 
desired to hold the voltage across the ter- 
minals of the telephone to a very low value, 
as it is voltage from line to line and not the 
voltage from line to ground which bums up 
the telephone coil. Just below this line-to-line 
gap are gaps from line to ground. 

An examination of a number of damaged 
telephones showed that in very many cases 
the end turns of the telephone bell coils were 
the ones that were damaged. Therefore two 
small choke coils were placed on the arrester, 
although at that time the idea was not to 
consider them as choke coils but simply as 
very highly insulated end turns of the tele- 
phone coils. Our tests have proved, however, 
that these coils also act very definitely as 
impedance coils when subjected to high 
frequency impulsi . 

This arrester has since been rearranged 
mechanically and a vacuum gap, man- 
ufactured by the General Electric Company, 
has been added in parallel with the air gap 
from line to line. The vacuum gaps that 
were used break down at approximately 
350 volts, thereby limiting the voltage 
across the telephone terminals to this low 
value. These vacuum gaps were first wired 
to the existing arresters and it was found 
that after they were added, practically com- 
plete protection was furnished the telephone. 

The arrester shown in Fig. 2 is one de- 
signed for lines which are not subjected to 
crosses with power lines of more than 

Fife. 2. Low Voltage Telephone Arrester 

2500 volts. In this arrester standard 
fuses are used and vacuum gaps are used 
entirely for relief gaps;, the two outside 
vacuum chambers being connected from 
each line to ground and the center -vacuum 
chamber connected from line to line. This 
arrester cannot be used where the operating 
voltage of the telephone line from line to 
ground is higher than about 50 volts. On 



ordinary telephone lines the voltage from 
lines to ground is much lower than this. 

Another arrester is being made for use 
where the voltage, in case of a cross with a 
power line, will be between 2.~i00 and 35,000 
volts. This arrester will be similar to the 
larger arrester in all electrical details but 
will be much smaller. 

The next arrester was made for use on 
telephone lines which were strung on the 
same towers and about ten and one-half 
feet from 1 10,000-volt power lines. This 
arrester is shown in front and side views in 
Figs. 3 and 4. The telephone lines are 
insulated for 22,000 volts and have a normal 
operating voltage to ground of approximately 
5.300 volts when drainage coils are discon- 
nected. The voltage from line to line is 
normally too low to be measured with com- 
mercial instruments. Fig. 5 shows the 
arrangement and connections of this arrester. 

Fig. 3. High Voltage Telephone Arrester 

Attention is called to the horn gaps shown 
at the top of this figure, which should pref- 
erably be mounted outside of the building, 
but between the telephone instrument and 
the first tower. This horn gap, which is set 
at about three-eighths of an inch, is a very 

essential part of the arrester, as it protects 
the top of the arrester frame and the fuses. 
and also the top insulator of the arrester. 
This protection is necessary, as an application 
of 50,000 volts or higher, continuously, on 

Fig. 4. High Voltage Arrester — Side View 

the top of the arrester will destroy this 
portion of it, although the arrester itself will 
afford the telephone instrument complete 
protection. It would be possible, of course, 
to build an arrester which would stand a 
continuous application of 110,000 volts with- 
out these auxiliary gaps, but the expense of 
providing 20-foot fuses mounted independently 
on 1 10,000-volt insulators is entirely out of 
the question when the same results can be 
achieved so simply. 

The arrester proper consists of what we 
call the gap unit, and expulsion fuses between 
the gap unit and the telephone line. The 
expulsion fuses are two feet long and are 
mounted on a very substantial frame which 
serves as a disconnecting switch. With the 
switch pulled the main part of the arrester 
is dead and fuses can be changed safely. 
The mechanical arrangement is very similar 
in principle to the 25,000-volt telephone 
arrester made by the General Electric 

The gap unit is mounted on a separate 
insulator and is tied with a plate to the bottom 
insulator of the fused switch. This stiffens 
the switch base and the gap unit base, making 
them both quite rigid. A number of materials 
were tried for the gap unit base before one 



was found which would stand the necessary 
high voltage test and also be sufficiently 
strong mechanically. Some samples of marble 
were found which were satisfactory, but 
more than half the bases made from selected 


- — Norn gaps, $g settmy, from each line to ground. 
Gaps mounted outside building. 

- — Samp fzpu/s.on fuses, Zft Zona. 

« — Cylinder gaps from each line to ground 
for tine near Sett/ng 

IIOOOO fotts Z Inches 

BSOOO • 06 ■■ 

/IOOO - 04- - 

Cylinder gap from tine to fine set . 0O4- inches. 

Choke coils wound *it h SO turns Mo 20 dec wire and 
cotton cord mound together cimitor to Parley coil, each 
layer insulated win three touens af ram/shed cambric 

1/acuum gap, from /me to tine, which breolfs down at 
3S0 volts 

Telephone transformer 25000 volt insulation 
bet wee n primary and secondary 

Fig. 5. High Voltage Arrester. Connections and Data 

marble had to be discarded because of par- 
tially conducting veins in the marble. The 
material finally used is a kind of Bakelite 

The gap unit consists of three brass cyl- 
inders connected as shown in Fig. 5, the two 
outside cylinders connecting to the telephone 
lines and the center one to ground. The 
spacing of these cylinders, which are adjust- 
able, should be such that the gaps between 
them will not arc over with the normal 
voltage of the telephone line, but should 
arc over at approximately 25 per cent higher 
voltage than normal. These gaps to ground, 
on account of their wide spacing (approx- 
imately 0.2 inch) offer practically no protec- 
tion to the telephone coils, but they do relieve 
the strains from line to ground which are 
impressed on the high voltage winding of the 
telephone transformer, and also act as a pro- 

tection against high voltage reaching the 
operator. Just below the ground gaps are two 
brass cylinders connected from line to line. 
These gaps, also adjustable, are set at 0.004 
inch and arc over at approximately 700 volts. 
In parallel with this air gap is a vacuum gap 
which breaks down at 350 volts. 

Choke coils, with the individual turns 
highly insulated, are mounted between the 
relief gaps and the telephone transformer. 
This telephone transformer, which has 25,000- 
volt insulation between the primary and 
secondary windings, is recommended in all 
cases on account of the protection it furnishes 
the operator. 

With the arrester connected we have not 
lost a single telephone transformer and only 
one telephone coil after several months 
service. Before the arresters were installed 
telephone transformers and telephones were 
burned out every few days during the light- 
ning season. 

Referring again to the expulsion fuses: 
these are fused with five-ampere fuse wire. 
This size was chosen because we did not 
wash to get a fuse so low that it would blow 
in case of a slight disturbance on the line; 
but on the other hand we did not wish a 
fuse so large that the gap units would be 
damaged after the fuses were blown repeatedly 
in sen-ice. Our operating results have shown 
that this size fuse is very satisfactory. 

Calibration curves were made of the arc- 
ing voltage between the particular cylinder 
gaps used on this arrester. It was found 
that the arcing voltage varied greatly with 
different kinds of knurling on the cylinders. 

When taking voltage measurements on 
the telephone line it was found that the 
voltage readings taken with a dynamometer 
voltmeter were very greatly in error. On 
the telephone line in question, the wave 
form of the voltage as shown by oscillograph 
records is very irregular, having a high peak. 
Therefore the voltage shown by the voltmeter 
was much lower than the voltage shown by- 
sphere gaps. As the spacing of the ground 
gaps should be in proportion to the voltage 
from the telephone line to ground, it is 
important that this point be noted as it 
caused us considerable trouble before we 
found out the reason for .the cylinder gaps 
discharging after they had been apparently 
set above the arcing voltage. 

High voltage may be applied to telephone 
lines in several different ways, all of which 
must be taken care of. The action of the 
arrester under different conditions of voltage 



application will be explained. If one tele- 
phone wire becomes crossed with one high 
voltage line, the current will flow over this 
wire and across the ground gap to ground 
without flowing through the telephone, pro- 
vided both ground gaps are set exactly the 
same. If it happens that the ground gap on 
the opposite side is set one or two thou- 
sandths of an inch closer than the gap on the 
wire carrying the high voltage, the current 
will tend to flow through the telephone 
instrument and discharge through the smaller 
gap. In this case the vacuum gap will come 
into action, shunting out the telephone and 
preventing damage to it. In addition to 
the current flowing to ground, another 
current must pass through the telephone 
or the arrester to charge the other line wire. 
This also is taken care of by the vacuum gap. 

If both telephone wires are crossed with 
one high voltage wire, the action of the 
arrester is practically the same, as the smallest 
ground gap will arc over first and the vacuum 
gap will discharge the other line without 
damage to the telephone. 

It is possible, although not probable, that 
each of the telephone wires will become 
crossed with separate power wires, thus 
impressing full line voltage across the tele- 
phone lines. In this case the vacuum gap 
will take the full discharge. 

The tremendous currents carried would 
destroy any piece of apparatus of reasonable 
size, if the cross continued for an appreciable 
length of time; therefore the five-ampere 
expulsion fuses are connected between the 
relief gaps and the line. In any of the cases 
above mentioned, or of a stroke of lightning 
on the line, these fuses clear up promptly. 
After the fuses have blown there is no further 
strain on the telephone or the arrester, but 
in case of a cross with the power line the 
tops of the fuses are still subjected to extreme 
voltage. The gaps outside the building will 
then arc over, relieving the stress at the top 
of the fuses until the telephone line burns 
down. This of course would take place 
anyway, because the insulators on the tele- 
phone line would arc over. Therefore these 
horn gaps will not cause any added trouble', 
but simply ensure that the protective appa- 
ratus is not damaged before the line does 
burn down. 

On tests made on the arrester it was 
found that if the voltage were raised slightly 
above the breakdown voltage of the vacuum 
gap. the vacuum gap would be destroyed 
after a time if the current were too low to 

blow the fuses and were held on continuously. 
It has been found that this condition is 
unusual and that the expense of renewing 
vacuum gaps has been negligible. After 
the vacuum gap has been destroyed, the 
cylinder gaps connected in parallel with it, 

Fig. 6. Telephone Arrester Operating on 118,000 
Volts from Power Line 

which are set at 0.004 inch, will arc over, 
thus preventing an extreme rise in voltage 
on the telephone. This cylinder gap breaks 
down at approximately 700 volts and will 
not entirely protect the telephone if the 
voltage is continued for a long time. None 
of these occurrences are necessary, however, 
if there is an operator in the station, as the 
telephone bell will continue to ring as long 
as voltage is applied. This should be a 
signal to the operator to clear the arrester 
by pulling out the fused switch. As the 
telephone is inoperative on account of exces- 
sive noise at this time, there is no objection 
to clearing the arrester from the line. 

Very extensive tests have been made on 
this arrester for the purpose of finding out if 
it would give complete protection to the 



telephone in cases of crosses with extremely 
high voltage lines. Tests were made in the 
General Electric Company's research labo- 
ratories with high voltage at 200,000 and 
.500,000 cycles. Other tests were made with 
power from high voltage 60-cvcle lines at 
22.000, 50,000 and 110,000 volts. 

Fig. 6 shows the arrester operating on 
118,000 volts connected directly on the 

that the cylinder gap which is in parallel 
with the vacuum gap did not discharge, 
but that the vacuum gap carried the full 
current. The expulsion fuses blew and the 
arc extinguished with a very sharp report, 
and immediately after the expulsion fuses 
cleared up the three-eighth-inch horn gap 
arced over. Then the lines connecting the 
horn gaps to the power line were fused. 

Fig. 7. Bank of Four High Voltage and Four Low Voltage Arresters at 
Boulevard Substation, Georgia Ry. & Pr. Co. 

power system from line to line. A standard 
General Electric telephone transformer and 
a telephone bell were connected to the .lower 
side of the arrester. When the test was 
completed this telephone transformer and 
bell were put back in service and are still 
operating, as they were not damaged. It 
will be noted that "three-eighth-inch horn gaps 
are connected two in series from line to 
line on the line side of the arrester, and that 
gaps are wired directly to the 110,000- 
volt power lines. Very close observation 
of the arrester at the time of this test showed 

This connection was made with 25-ampere 
fuse wire, as we did not wish to subject the 
110,000-volt power system to a continued 
short circuit. 

This test, as well as a number of others, 
showed that the vacuum gap will apparently 
take care of an enormous current without 
damage to itself, provided the current is 
interrupted promptly, as is done by these 
expulsion fuses. Before this extreme test 
was made, the breakdown voltage of the 
vacuum gap was 350, after the test the 
breakdown voltage was 390. 



By C. N. Moore 
Research Laboratory, General Electric Company 

Until very recently the thought of an X-ray tube immediately called to mind its application to medicine 
and surgery. The Coolidge tube, however, has broadened the useful scope of the application of X-rays so 
successfully that it is now employed widely for engineering purposes. In recent issues of the General 
Electric Review we have described the method of examining steel castings and copper castings for internal 
defects; and the present article treats of the X-ray inspection of built-up mica. — Editor. 

The process of manufacturing "built-up" 
mica for use as an insulating material in 
electrical machinery consists essentially in 
pasting together, at an elevated temperature 
under pressure, thin flakes of mica with a 
suitable binder, planing down the resulting 
product to the required thickness, and 
cutting it into sheets of the required size. 
In this process, certain defects which would 
affect the insulating qualities of the finished 
product have to be guarded against. Among 
these are the presence of foreign materials 
of a metallic nature, and of areas not of the 
required thickness. In practice, these defects 
are detected by subjecting the material to 
very careful visual inspection and 
gauging with a micrometer. This, 
however, entails considerable labor. 
The successful application of X-rays 
to the detecting of defects in such 
materials as steel and copper cast- 
ings, already described in earlier 
issues of this publication, suggested 
the possibility of utilizing X-rays as 
a means of increasing the efficiency 
of the regular inspection of mica. 

With this end in view, Dr. 
Davey and the writer obtained 
micas (some known to be good 
and others known to be defective) 
for examination in the Research 
Laboratory. These samples 
were about 0.032 of an inch in 
thickness and had been cut into 
small sheets of the required size for 
placing in the commutators. These 
pieces were placed upon a fluores- 
cent screen in a specially designed 
viewing box (Fig. 1) at a distance 
of 20 inches from a Coolidge X-ray 
tube. When the tube was operating with a 
current of about 6 milli-amperes and a parallel 
spark gap of six inches, the structure of the 
mica, as shown on the fluorescent screen, could 
be viewed from the outside of the box by means 
of a mirror set at an angle of 45 deg. to the 

Some of the samples examined contained 
small particles of iron oxide not visible to 
the eye on the surface of the sheet of mica. 
As iron oxide is much more opaque than 
mica to X-rays, this material showed up as 
black spots in the image of the mica on the 
fluorescent screen. Other samples examined 
contained small sections not as thick as the 
main portion of the sheet. • These sections, 
being more transparent to X-rays, showed up 
as light spots in the image on the screen. 
Samples of uniform thickness which con- 
tained no foreign material gave images of 
uniform density upon the screen. It was 
found that the examination could be made 

/v£D W& vV'/vc 

Fi 8 

1 . Arrangement of Coolidge Tube and Viewing Box for 
Inspection of Built-up Mica 

very accurately and rapidly, one glance at 
the" image on the screen being sufficient to 
detect the presence of any defects. 

The nature of the images on the fluorescent 
screen is shown in the radiographs. These 
were taken on Seed X-ray plates, with an 
exposure of five minutes at a distance of 





OTHB *1m * •- , *. 



£ S 
in „ 


.- y 

O tJ 

X £ 




















30 inches from a Coolidge tube. The tube 
was operated from an induction coil on 10 
milli-amperes with a parallel spark gap of 
four inches. Fig. 2 shows the radiograph 
of three sheets of fairly uniform thickness. 
The various flakes of mica which go together 
to make up the finished sheet are plainly 
visible. As these flakes are in most cases only 
a few thousandths of an inch in thickness, 
this radiograph shows what small differences 
of thickness may be detected by means of the 
X-rays and the fluorescent screen. Fig. 3 illus- 
trates this more clearly. In this case a sheet of 
mica 0.050 of an inch thick was p'laned down so 
that successive sections were 0.045, 0.035 and 
0.020 inch thick. The radiograph of this sheet 
shows that a difference in thickness of 0.005 of 
an inch may readily be detected. 

The ease with which foreign material 
may be detected is shown by Fig. 4. The 
particles of iron oxide present in this par- 
ticular case were not visible on the sur- 
face, but they are plainly visible as black 
spots in the radiograph taken of the sheets 
of mica. 

Fig. 5 shows a radiograph of four sheets of 
mica 0.032 of an inch thick with small areas 
considerably thinner than the main portion 
of the sheet. These thinner areas show up 
as light spots in the radiograph. 

The results obtained on an experimental 
scale in the laboratory have demonstrated 
the adaptability of the X-ray apparatus as 
a factory tool for the inspection not only of 
"built-up" mica but of any similar material 
of not too great a thickness. 


By W. E. Rvder 

Research Laboratory, General Electric Company 

A number of years ago it was universally believed that the purity of a piece of iron was a direct indication 
of the serviceability of the magnetic properties of that sample. Later experiment has disproved this belief 
and showed that the changes produced in the magnetic characteristics of iron by the addition of certain 
foreign materials is, in reality, commercially beneficial. The following article discusses the effects upon tin- 
magnetic properties of iron by the addition of silicon, aluminum, arsenic, tin, copper, cobalt, nickel, chro- 
mium, tungsten, molybdenum, sulphur, phosphorus, and oxygen. It also discusses the non-ferrous alloys and 
makes reference to several prominent theories purporting to explain the phenomenon of magnetism. — Editor. 

There is a legend that 24 centuries B.C. 
Hoang Ti, Imperial navigator for China, 
piloted his fleet of junks to victory by means 
of a floating piece of loadstone. 

It was not until the time of Marco Polo, 
however, that its use as a compass was known 
in Europe. Frequent mention of its peculiar 
properties were made before this by Lucretius, 
Pliny and Plato, and it is said that the 
Priests of Samothrace made a steady and 
comfortable income from the sale of magnet- 
ized iron rings which were supposed to cure 
all manner of ills. Thus was born the idea 
which we now have expressed in the modern 
' ' electric belt, ' ' and magnetism has been a most 
lucrative field for all kinds of medical quacks 
down to the present day. 

The two metals most used in the electrical 
industry of today are copper and iron, and 
any saving, even in the smallest amount, of 
either, in electrical design, means an immense 
saving in the total quantity used. In iron, 
two diametrically opposite sets of properties 
are desired depending upon the uses to which 
it is to be put. For permanent magnets, it 

is desirable to have a high coercive force and 
a high remanence; while for magnetic cir- 
cuits a high permeability combined with low 
coercive force is most desirable. The latter 
of these two uses is by far the most important 
in that it involves considerably more material 
and all kinds of electrical generating ap- 

Besides the desirability of having a high 
permeability and a low watt loss, it is also 
necessary to have permanency of magnetic 
quality, i.e., the losses as calculated in the 
design must not change under the conditions 
in which the apparatus is operated. A high 
electrical resistivity is also desirable so that 
the Foucault currents may be limited. 

For a long time pure iron was considered 
the best possible material for magnetic cir- 
cuits, and specifications always called for a 
pure grade of Norway iron — then, the 
purest commercial grade of iron. This 
material satisfied the demands for a fairly 
good permeability and low hysteresis loss, 
but unfortunately after running for a short 
time it was found that both the permeability 



and hysteresis had deteriorated, often in the 
case of the latter, as much as 100 per cent or 
more. Several years ago some engineers 
discovered that slight amounts of impurity, 
such as silicon and manganese, did not injure 


b - Coercive - Porce 










as y.o is 
Percent C. 



1.0 AS 

Percent C 


Fig. 1. Effect of Carbon Content upon Resistivity and Coercive 
Force. ^Gumlich) 

the magnetic properties and did prevent to a 
large extent the deterioration of magnetic 
quality with time. These impurities are of 
the same kind, and are about the same in 
amount, as exist in the best grade of basic 
open hearth steel. This represented the first 
great step forward, and in 1900, Barrett, 
Brown and Hadfield published their results 
obtained on alloys of iron with as high as 
five per cent of silicon or aluminum which 
showed the remarkable fact that additions 
of non-metals or semi-metals, which had no 
magnetic properties in themselves, would 
still improve the magnetic quality of iron. 
Since that time there have been several 
improvements, but mostly in the mechanical 
working of the sheets and in their annealing 
and heat treatment. The useful magnetic 
properties of iron and its alloys are effected 
in three ways; first, by composition, second, 
by mechanical treatment during the process 
of manufacture, and third by the crystalline 

This article will deal only with the effect 
of chemical composition upon the magnetic 
properties of steel. 

Carbon, always present to a greater or less 
extent in commercial steels, has a decided 
influence upon their magnetic properties. 
Gumlich, working at the Physikalisch-Tech- 
nische Reichsanstalt, recently published some 
interesting results upon the influence of this 
element. These are given in part below. 

The electrical resistance rises about 0.06 
ohm per m. and sq. mm. for each per cent of 
C. The curve, however, bends at about one 
per cent C. and above that the increase is less 
rapid. Practically the same result is obtained 
when the coercive force is plotted against 
per cent carbon. 

We know that in all slowly cooled iron- 
carbon alloys the carbon exists as pearlite 
up to the eutectoid point, i.e., about 0.85 
per cent C, above this percentage the carbon 
exists as cementite (FeaC), the normal carbide 
of iron, and the curve indicates that the 
cementite carbon diminishes the conductivity 
less than does the pearlite carbon. This is 
explainable by the lamellary structure of the 
pearlite which is made up of alternate layers 
of FesC and pure iron each about 1/25000 of 
an inch or less in thickness. 

The effect upon the hysteresis loop is to 
make it broader and lower because the per- 
meability is decreased from about \i max = 
5000 for 0.02 per cent C. to y. max. =450 for 
1.8 per cent C. 


^ /SOOO 

k /eooo 

* /?000 

O OS 1.0 15 2.0 2S 30 SS 40 

Percent C. 

Fig. 2. Showing Decrease in Saturation Value for Increased 

Carbon Content in Annealed and Hardened 

Steels. (Gumlich) 

The value of saturation, 4 it I, which is the 
true index for magnetic quality, diminishes 
about 1400 for each per cent of carbon present, 
so that for pearlite (0.86 per cent C.) its value 
is 20,200 and for cementite it can be calculated 







+ \ 








l .uns 




v a/ 




J ot 






to be about 12,500, considering pure iron to 
have a value of 21,600 (see curve, Fig. 2). 
For hardened steels it will be observed that 
the decrease is more rapid owing to the per- 
centage of C. held in solution. The break at 
1.4 per cent is due to the fact that at the 
quenching temperature, 850 deg. C, no more 
carbon was dissolved so the curve becomes 
parallel to that for the annealed samples which 
have no dissolved C. 

If, now, these alloys be subjected to harden- 
ing at different temperatures, we find that 
the coercive force and resistance rise directly 
in proportion to the percentage of dissolved 
carbon as is shown in the curve (Fig. 3). 

Only the curve for specimens quenched 
at 800 deg. C. are given and it will be observed 
that there is again a sharp break at about 
one per cent C. This is due to the fact that 
only this percentage of carbon is soluble in 
the iron at this temperature. At higher 
temperatures the curve is smooth, though 
it bends a little due, no doubt, to the forma- 
tion of austenite at the higher quenching 

Silicon and Aluminum 

It has been mentioned before that silicon 
has a decided effect upon the magnetic quality. 
It is difficult to understand at first how so 















X+- at SfO° 

K 1+ 



o as /.o /.£ 2.0 a<5 3.0 
Per cent C 

Fig. 3. Effect of Hardening upon the Electric Resistance of 
Carbon Alloys. (Gumlich) 

magnetically inert an element as silicon could 
produce a better magnetic quality. The 
term "better," however, is used in the sense 
of{more useful, rather than in an "absolute" 
sense, for the improvement obtained from 

silicon is not a direct one; first, because the 
magnetic quality does not improve in pro- 
portion to the increasing silicon content, and 
second, because the saturation value falls off 
steadily with increased percentage of silicon 




% /£>000 



£ /eooo 













3 4-^6 
Per cent Si. 


Fig. 4 Showing Decrease in Saturation Value of_ Steels with 
Increased Si Content. ^Gumlich) 

from 21,600 to 16,500 for 8.5 per cent silicon, 
as shown in the curve, Fig. 4. 

This curve shows that the silicon acts 
largely as a foreign substance, diminishing 
the active cross-section of the iron and there- 
fore the saturation value. 

In brief, the real effect of the silicon upon 
the steel is this: 

(1 ) It prevents the formation of hardening 
carbon, even with comparatively quick cool- 
ing, and with the higher percentage (3-5) 
even the formation of pearlite is prevented 
and all the carbon exists in the harmless 
graphitic state. 

Charpy and Cornu (C.R., May 26, 1913), 
show that at least 800 deg. must be attained 
in annealing to change all of the pearlite to 
graphite for steels having a silicon content 
of 3.8 per cent, and in general the temperature 
at which the separation begins is lower as the 
silicon content is higher. 

(2) It cleanses the metal of harmful oxides 
and dissolved gases. 

(3) It produces a larger grain structure in 
the metal. 



(4) It increases the resistivity of the 
metal from 12 to about 60 or 75 microhms per 
cm. cube, depending upon the contents. 
(See curve, Fig. 5.) 

O / 2 3 4~ & 
Per cent Si 

Fig. 5. Effect of Si upon Resistivity. (Burgess) 

Paglianti (Metallurgist 9, pp. 217-230) 
made up a series of Si alloys containing 0.2 
to 5 per cent Si and drew the following con- 
clusions: The specific gravity diminishes 
regularly from 7.87 to 7.57 with increased Si 
over the range studied. For high induction 
the permeability falls off with increased Si, 
but for B = 7000 there is a considerable en- 
hancement of the quality with the addition of 
Si. With annealed alloys the maximum 
value of n was 1500 with 0.2 per cent Si, 
3300 with 2 per cent Si, and 2800 with 5 
per cent Si. For B = 13,000 the hysteresis 
loss with 2 per cent to 5 per cent Si was 
only half that with 0.2 per cent Si. 

The results obtained by Burgess (Met. 
Client. Eng. 8, 131) are shown in the curves, 
Fig. 6. 

The results which have been obtained in the 
laboratory with ring samples of 0.015-inch 
sheet differ somewhat from each of these and 
are shown in Fig. 7. 

Silicon has the effect, however, of making 
the sheets more brittle. Brinell hardness 
numbers increase fairly uniformly from 125 
with 0.2 per cent Si to 290 with 5 per cent Si 
(Paglianti). Its use is therefore limited to 
stationary apparatus such as transformers, 
although some alloys with 1 to 2J4 per 
cent are used in induction motors. In general, 
however, generators and motors operate at 

such high flux density that the advantage of 
low watt loss is offset by the decrease in per- 
meability at these high densities, so it is still 
the practice in this kind of apparatus to use 
a pure grade of open hearth steel with only 
about 0.02 to 0.1 per cent Si to limit the 

In general, aluminum has the same effect 
upon the magnetic properties as silicon. It is 
not, however, in as general commercial use as 

Arsenic and Tin 

Burgess and Aston have shown that these 
elements, like silicon and aluminum, also have 
the effect of raising the electrical resistance 
of the steel and of inducing a large grain 
structure which has a good effect upon mag- 
netic hysteresis. 

The effect of tin is to increase the permeabil- 
ity of higher ranges and to decrease the hyster- 
esis loss to a lower value even than silicon. 
This loss decreases gradually with increasing 
percentage content of tin. 

Tin is, in this respect, better than arsenic 
which also reduces the hysteresis loss more 
than silicon when as much as 3.5 per cent is 
added, but it has the disadvantage of being 
verv brittle. 









ZO 40 60 80 IOO /20 MO /60 /80 200 

Fig. 6. Magnetization Curves. Iron-Silicon Alloys. Annealed 
675 Deg. (Burgess) 

Both arsenic and tin have the advantage 
over silicon of increasing the permeability 
in the higher working ranges of density. 
This increase is hard to explain in view of 
the fact that, like silicon, they are both non- 

F— <r. 




















] YJt 

1 1// 









magnetic elements, and more accurate work 
on these alloys is necessary before we can 
explain their results. The action is, no doubt, 
in view of the more recent work on gases in 
pure iron, one of a cleansing nature. 


The addition of copper has been suggested 
as a means of increasing the permeability of 
iron. In fact, a patent has been taken out upon 
such a process. The writer has been able to 
find no case, however, in which the permea- 
bility was increased or the coercive force de- 
creased. In fact, the contrary was found 
to be true, the magnetic quality deteriorating 
almost in proportion to the percentage of 
copper added. The one advantage to be 
gained is that of increased tensile strength in 
1 per cent to 2 per cent copper alloys 
without a very great decrease in magnetic 
quality. The electrical resistance also rises 
to a maximum of 17 microhms for 1^ per 
cent copper at which point it is about 1.40 
times that of standard iron. 

Cobalt and Nickel 

The writer knows of no published results 
on a systematic study of cobalt alloys with 
iron, and such results would undoubtedly 
be quite instructive in view of the interesting 
results obtained from nickel. Weiss has 
found, however, that the alloy Fe2Co has a 
saturation value 10 per cent higher than pure 
iron. This is the only alloy known having 
such properties. The writer's own results 
have checked this finding. 

Nickel (Burgess & Aston, Met. Chem. Eng. 
8, pp. 23-26), when added in quantities less 
than 2 per cent, causes little change in 
magnetic quality. With increasing nickel 
content, the permeability rapidly falls off and 
the alloys of 25-30 per cent of nickel are almost 
completely non-magnetic. 

A still further addition of nickel again 
improves the quality up to 50 per cent, so 
that for fields under 14,000 B the permea- 
bility is higher than that of pure iron and its 
hysteresis loss is only 50 per cent of that of 
iron. The sudden drop in permeability above 
14,000 B, however, together with its cost 
prohibits its use as a transformer material. 

Chromium, Tungsten, Molybdenum, etc. 

Metals, such as chromium, tungsten, molyb- 
denum and manganese, have the general 
property of increasing magnetic hardness, 
that is, they increase the remanence and more 
particularly the coercive force. These are the 
properties which are desirable for permanent 

magnets. It has been found that they work 
best in combination with carbon or some other 
element, such as silicon or vanadium. Nickel, 
though it sometimes aids, in general is found 
to be injurious to permanent magnetic 




/600 2400 3200 4000 4600 J600 6400 








^3.7 & 



s. "■ 


























O /Q SO 30 40 SO 60 70 80 SO /OO I/O SO /30/40I50I60 

Fig. 7. Effect of Silicon upon the Permeability and Magnetiza- 
tion Curves 

quality. Moir (Phil. Mag., May, 1914), has 
made a study of the magnetic properties of a 
graded series of chromium alloys. 

The curve (Fig. 8) shows the considerable 
decrease in permeability caused by the addi- 
tion of 25 per cent of chromium, both before 
and after annealing. 

Steels which are practically non-magnetic 
can be made by the addition of 12 per cent 
Mn with about 1 per cent C. which is known 
as Hadfield's manganese steel. This material 
has a practically constant permeability of 
about 1.3. The manganese may be increased 
to 18 per cent in some cases with the same 
result. A slight change in heat treatment may 
create or destroy the magnetic property in 
such a steel, because Mn has the property of 
lowering the change point so much that the 
gamma iron remains unchanged with reason- 
ably quick cooling and gamma iron is non- 

Sulphur, Phosphorus and Oxygen 

Of the elements, sulphur, phosphorus and 
oxygen, it may be said in general that they 



are all injurious. Even though they exist 
in most cases in very small percentages, they 
combine in such a manner with the iron to 
form sulphides of iron and manganese, 
phosphides and oxides, that they occupy 





400 £00 






z ff± 



L /5-A- 






Magnetizotion ondPermeobj/iiy Curves 
BS^o ferro Chrome 
O.Bfr C 
Afoj i/nonneofea 'rod 
ffQ24/?r7eo,'ea'i&00 a C m vocuum. 
3urrows Method 

/5 8000 































no /ao 130 iic 

Fig. 8. Magnetization and Permeability Curves; 25 per cent Ferro-chrome, 

0.2 per cent carbon. No. 1, Unannealed Rod; No. 2. Annealed 

1000 Deg. C. in Vacuum, Burrows Method. 

considerably more space than the analysis 

In any case their removal always results 
in a considerable increase in magnetic quality. 

This brings us to a consideration of some 
of the more recent work on pure electrolytic 
iron. As pure as commercial iron is, the 
further purification by electrolysis has made 
it possible to produce material having a very 
high permeability and low hysteresis loss. 
Dr. Breslauer in an article in the E.T.Z. has 
calculated that a saving of from 15 to 50 per 
cent in the weight of iron is possible. 

The chief disadvantage of this material is 
its low resistivity — 10 microhms per cm- 3 . 
In generators, however, its high permea- 
bility more than offsets its high eddy loss, 
which is of less consequence in this form of 

In spite of the high purity of this material, 
a still greater improvement has been obtained 
by heating the sheets of alloyed or pure iron 
to a high temperature in vacuo, or by fusing 
in vacuo. The effect of this treatment is evi- 
dently to remove practically the last traces of 
impurity, notably sulphur and oxygen. 

The important point about this work on 
extremely pure iron is that it shows that the 
good results obtained by additions of other 
elements are all secondary, i.e., their effect 
is one of removing material that is more 

injurious, magnetically, than the metal added. 
The additional effects are those of increased 
resistivity — always at the expense of magnetic 
quality however — and an increase in the size 
of the grain. 

Non-Ferrous Alloys 

About the non-ferrous 
magnetic alloys much is 
yet to be learned. These 
alloys are named after 
Dr. Heusler who first 
called attention to them 
in 1898. A clear under- 
standing of the reason for 
their magnetic properties 
and the laws governing 
them would add much to 
our understanding of the 
ultimate nature of magnet- 
ism and a brief word here 
as to their composition and 
properties would not be 
out of place. They consist 
approximately of 62.5 per 
cent Cu, 23.5 Mn and 14 
Al, i.e., Mn and Al in the 
proportion of their respec- 
tive atomic weights. Their magnetic values 
are very low compared with good iron; the 
best value of permeability being about 90-100 
at a density of 2200 B. A magnetizing force 
of 70 which gives a flux of about 17,000 B 
(lines per sq. cm.) in soft iron gives only 4000 
B in the best of these alloys. 

The likeness in structure between these 
alloys and iron is striking, but no definite 
structure or kind of cryst al has been attachable 
to the strongly magnetic forms as differing 
from the weak or non-magnetic modifications. 
Prof. A. A. Knowlton describes three kinds 
of crystals separated by their different be- 
havior under the action of the etching re- 
agent, and finds a definite relation between 
one of these and the saturation value of I, 
but subsequent workers have not been able 
to find this relation or to isolate this particular 
crystal structure. The following statement 
is taken from a paper by Heusler and Take 
(Trans. Faraday Society, October, 1912): 

1. In order to account for the pronounced 
ferro-magnetism of the Heusler alloys, aluminum 
or tin-manganese bronzes, Guillaume assumes with 
Faraday that pure manganese exists in a strongly 
magnetic modification, which undergoes transforma- 
tion at a very low temperature; this temperature 
is said to be raised by the addition of Al or Sn, so 
that the magnetism of these alloys becomes appar- 
ent. This hypothesis is not based upon facts, and 
the arguments are not sufficiently supported; it 
would, moreover, not explain the strong ferro- 



magnetism of the Heusler manganese alloys with 
As, Sb, Bi and B. 

2. Heusler, on the other hand, has advanced a 
hypothesis which explains all the phenomena; for 
he has not only discovered the ferro-magnetie 
alloys, but he also first proved, in 1903, that the 
appearance of the strong ferromagnetism is due to 
the formation of chemical compounds, and that 
the ferromagnetism is thus a molecular phenomenon. 

This article may appear more or less a 
jumble of facts having, in many cases, no 
continuity of reason. This is largely due to 
the fact that no theory has yet been evolved 
which fits in with all of the facts, and research 
along magnetic lines is largely a matter of 
cut and try, especially when dealing with 
alloys of a possible commercial importance. 
Several theories have been evolved and they 
are all ingenious, but they have always 
followed, rather than led, the experimental 
work. The latest and most interesting theory 
has been the adaptation, by Langevin of 
Paris, of the electron theory to magnetism. 
It does not attempt to predict, of course, 
what element is to be added to iron to make it 
more useful, or what element is the most 
harmful, but these things must be left to be 
decided by experiment. 

As far as a practical use can be made of it, 
Langevin's theory scarcely goes much further 
than Ampere's resistanceless circuit theory or 
the well known molecular theory of Ewing. 
From a mathematical and theoretical point 
of view it undoubtedly adds much toward 
correlating our recent views of energy and 
matter. (See Dushman, General Electric 
Review, 1914, July, Sept., Oct. and Dec.) 

The recent work of Bragg (Phil. Mag., 
Sept., 1914), who has succeeded in determin- 
ing not only the crystalline structure but the 
relative location of different atoms in a mole- 
cule, has opened the way for an explanation 
of the ultimate nature of magnetism which will 
take into consideration the relation of the 

position of the elements in the crystalline 
space lattice to the resulting external effect 
which we call magnetism. After a thorough 
study of iron and its alloys by this method 
we shall be able to completely correlate the 
jumble of facts which now constitutes our 
knowledge of the effect of added elements 
upon the magnetic properties of iron. The 
magnetic and non-magnetic phases of the 
Heusler alloys, or Hadfield's manganese steels, 
may then be explained by a deformation, or 
permanent alteration of the atoms from their 
original equilibrium in their crystalline space 
lattice. Magnetic hardening which so often 
accompanies real hardening may then be 
explained on the same ground, viz., that of a 
strain set up between the atoms in their space 
relation in the crystal. 

The chief difficulty encountered by the 
experimenter has heretofore been that of 
obtaining accurate magnetic tests. Most 
workers along this line have used rods or 
wires for convenience, and have obtained 
results which, while they have a comparative 
value, are quite inaccurate and misleading. 
Burgess and Aston, who have covered more 
ground than any one else in the investiga- 
tion of alloys for magnetic quality, used 
this form of test piece, recognizing its limita- 

The ring method has long been the standard 
of magnetic testing. C. W. Burrows (Bull. 
Bur. Stds., 6) has developed an end compen- 
sation method for straight bars which makes 
it now possible to obtain permeability results 
on bars which are as reliable as those obtained 
from rings. 

This article has only dealt with the effects 
of composition upon magnetic quality. The 
mechanical treatment and crystalline struc- 
ture also have a very decided influence which 
can, in extreme conditions, wipe out all the 
beneficial results of a proper composition. 


Part II 
By J. P. Minton 

Research Laboratory, Pittsfield Works, General Electric Company 

In our previous article of this series the author dealt with the properties of cathode rays and electrons. 
In the present installation the principles of the conservation of electricity and the nature of its flow are 
considered first; followed by a discussion of the electron theory of matter and electricity and its application 
to the electrical and thermal conductivity through metals, and also the relationship between these two phe- 
nomena. A modified electron theory is also discussed, and the article concluded with a brief summary of the 
subjects covered in this contribution. — Editor. 



In the first article, on the properties of 
cathode rays and electrons, we considered the 
experimental results upon which the elec- 
tron theory is based. No conclusions were 
drawn, whatsoever, except those which could 
be drawn from the data at hand. In this 
article, however, the electron theory of matter 
and electricity will be briefly developed, and 
applied to the conduction of electricity 
through metals. Some contradictions will be 
encountered, but this means that the nature 
and the dynamical behavior of a complex 
piece of matter is certainly not fully known. 
and does not mean that we are on the wrong 
path. On the contrary, it is believed that we 
are, indeed, on the right path to the real 
understanding of matter and electricity, even 
though this understanding may be far dis- 

The subjects to be considered in this article 

I. Preliminary statements on the Prin- 
ciple of the Conservation of Elec- 
tricity, and on the nature of Flow 
of Electricity. 
II. Electron Theory of Matter and 
III. Application of this theory to: 

(a) Electrical Conductivitv through 


(b) Thermal Conductivitv through 

IV Relation between Thermal and Elec- 
trical Conductivities of Metals. 
V. Number of Free Electrons in a 
Cubic Centimeter of Metal. 
VI. Modified Electron Theory, and its 
Application to Electrical and Ther- 
mal Conductivities of Metals. 
VII Summary and Conclusions. 
We shall now take up these various subjects 
in the order riven. 


Electricity, like matter and energy, is inde- 
structible 'and cannot be created. If a certain 
amount of electricity disappears from one 
place, it always appears in another. For 
example, if a positively charged sphere is 
connected to one which is not charged, there 
is a flow of electricity from the former to the 
latter. The quantity of electricity lost by 
one is exactly equal to that gained by the 
other. Furthermore, the apparent destruc- 
tion of electricity is due to the neutralizing 
effect of positive and negative charges on 
each other. 

Again, a positively charged body will in- 
duce a negative charge on a body placed near 
it. At first thought one might say that 
electricity has been created in this case. 
Such is not the case, however, for the posi- 
tively charged body simply attracts the nega- 
tive constituents and repels the positive ones 
of the uncharged material. Hence, electricity 
is separated, not created, and this separation 
continues until equilibrium exists between the 
forces thus set up. Hence, there is no such a 
thing as a generator of electricity, but should 
properly be called an electrical separator. 
Consequently, any theory of electricity must 
explain how this separation of the already 
existing charges is brought about. We shall 
see that the electron theory of electricity- 
does this very nicely. 

Since we are to consider metallic conduction 
in this paper, it will be important to first 
point out how electricity is carried from one 
place to another. There are displacement 
(Id), convection (I c ), and conduction (I c ) 
currents, and the total current (I) passing 
through any medium is, I = (Jd+Iv+Ic)- 

In regard to displacement currents we can 
imagine an air bubble within a piece of soft 
rubber. Now. if the air bubble expands or 



contracts, the rubber will be displaced 
throughout its volume, and this corresponds 
to displacement currents, as in a perfect 
vacuum condenser where there are only dis- 
placement currents. For an alternating 
potential there is a corresponding displace- 
ment current, but for a direct potential the 
displacement current becomes zero when the 
stress in the medium balances the force pro- 
ducing the strain. 

The flow of ions through electrolytes and 
gases, of charged particles of mercury vapor 
in mercury lamps and rectifiers, of electrons 
in gases and solids, correspond to what are 
called convection currents. The electron 
theory deals entirely with these currents. 

Now, as to the nature of conduction current 
we know nothing, and do not even know that 
there is such a thing. Indeed, in the light of 
our present knowledge of the electron theory 
of metallic conduction, it appears that there 
is no such a thing as a conduction current in 
metals, but that it is all convection current 
due to the electrons. Let us, then, take up 
the development of the electron theory and 
its application to metallic conduction. 


We have seen in the previous paper that 
electrons come from metallic cathodes and 
they must, therefore, exist in metals, being 
torn out of them by the electrical forces 
brought into action. We also saw that when 
the cathode rays strike a fluorescent sub- 
stance, say calcium tungstate, some of the elec- 
trons remain in the substance. There are many 
other illustrations which we could use to show 
that electrons form an important part in the 
formation of a piece of matter. It has been 
shown, as in the case of cathode rays, that 
electrons exist free from matter. Now, the 
question is how do the}' combine with matter ' 
Since in gases there are neutral molecules and 
atoms, it follows that electrons must exist 
within these in order that the positive and 
negative electricity may balance each other. 
So when gases are solidified, electrons must 
form an important part of the resulting solid 
matter being bound within the molecules 
and atoms. We have ample evidence to 
prove the free electrons exist in gases, and 
therefore solids if the gases are solidified. 
Similar remarks apply to vapors, such as 
mercury vapor for example. Electrons are 
shot off from radio-active substances and 
metals heated to a high temperature. In the 
same way we can show the electrons exist in 

all substances. Since, in many cases, they 
can be liberated from these substances with- 
out apparent disintegration of the material, 
it follows that these electrons must exist in 
the free as well as the combined states in all 
substances. The free state refers to their 
existence outside the atoms or molecules, and 
the combined state means that they exist 
within these. 

Having shown that electrons exist in all 
substances, let us see to what extent they are 
in the free and combined states in these 
substances. Now, all electronegative ele- 
ments, like chlorine, sulphur, etc., take on a 
negative charge, and hence most of the 
electrons are in the combined state within 
these. In the case of electropositive ele- 
ments, like silver, copper, zinc, etc., we have 
just the opposite state of affairs, and in these 
there are a great many free electrons as com- 
pared with those in the electronegative ele- 
ments. This means that there is always 
negative electricity in a piece of matter, and 
there must, therefore, always be positive 
electricity. Take the case of a salt, say silver 
chloride, as an example. The silver is electro- 
positive while the chlorine part of the molecule 
is electronegative. An atom being electro- 
positive means, from the theory we are 
developing, that the atom has liberated some 
of its electrons. Hence this leaves the silver 
part of the molecule positive, and, therefore, 
it is difficult for the silver atom to lose more 
electrons because of the electrical forces 
brought to bear as a result of the loss of 
electrons. Since the chlorine part of the 
molecule is electronegative (that is, it has a 
capacity for taking on electrons to give it a 
negative charge) it means that the electrons 
liberated by the silver atom are taken up by 
the chlorine radical. After this radical has 
taken on some electrons, its negative charge 
will tend to oppose the taking on of any more 
of them. Both of the actions here described 
continue until equilibrium is established, when 
the exchange of electrons ceases. We natu- 
rally expect, therefore, that most of the elec- 
trons in a salt are in the bound or combined 
state. However, it is quite likely that a small 
percentage of all the electrons present are in 
the free state. This same line of reasoning 
applies to acids and bases. The conception 
of free and bound electrons will be utilized 
in applying the electron theory. 

Our conclusions are based upon our ex- 
periences, and our experiences indicate that a 
piece of matter is built up of positive and 
negative electricity, electrons, atoms, mole- 



cules, bound and free energy, and things that 
we do not yet know about. So, a piece of 
matter is an extremely complicated thing. 
Let us, now. take up the electron theory of 
the conduction of electricity through metals. 


From the above discussion we see that there 
are a great many free electrons in a piece of 
metal as well as combined ones. Now, the 
first theory we shall develop considers only 
the free electrons and that these are in a 
continual state of vibration between the 
metallic molecules and atoms. This movement 
is in all conceivable directions, so that there 
is no resultant transfer of electrons along a 
wire. In this case there is no current flowing. 
Suppose, however, an electric force (2s) is 
applied to the wire. Then, there will be a 
resultant transfer of electrons along the wire 
in the direction opposite to the electric force 
on account of the electrons possessing a nega- 
tive charge. 

We see, therefore, that we are dealing with 
two electronic velocities; the first (v) being 
that due to the random vibration between the 
collisions of the electrons with one another and 
with the molecules and atoms, and the second 
(U) being the resultant drift of the electrons 
along a wire constituting the flow of current ; 
(U) and (v) are average values, of course. 
Xow, in this theory we apply Boltzman's 
principle, namely, that in any gas the mean 
kinetic energy of all molecules are equal. So, 
we assume that the free electrons of a piece 
of metal act like a gas, and hence, the kinetic 
energy of the electron must be the same as 
that of a hydrogen molecule at the same 
temperature. Since the mass of the electron 

is about — of that of the hydrogen molecule, 

it follows that (tr) must be about 3500 times 
the square of the velocity of the hydrogen 
molecule. The mean velocity of the latter, 
as shown by the kinetic theory of gases, at 
deg. C. is about 1.7 X10 5 cm. sec. so that v 2 = 
<(1.7X10 5 ) 2 , or i' = 2X10 7 cm. sec. 
approximately, which is the velocity (y) of 
the electron. (U) is so much smaller than 
this that it need not be considered in obtain- 
ing the mean kinetic energy of the electron 
within the wire under consideration. For, 
suppose [7 = 1 cm. sec, then from equation 
(1) below I = Xe. Now, we shall see later 
in this article that .V= 10 24 approximatelv, so 

tbat r. '°"X4.8x 1 o-.- xl0orJ , 10>amp 

approximately, which is very large. Since 
this current is excessive, it follows that (U) 
must be small. 

Let us, now, derive expressions for the mag- 
nitudes of the current and the electric con- 
ductance from this theory. We all know that 
the current (I) flowing across an area of 
one sq. cm. equals the charge per cu. cm. 
multiplied by the velocity of drift (U). If 
there are (A r ) free electrons per cu. cm. and 
e is the charge on each electron, then Ne 
equal the charge per cu. cm. So that. 

I=NeU. (1) 

We can get an expression for (U) in the 
following manner. During the time an elec- 
tron is moving along its mean free path it 
traverses a distance due to (E) : 

D = 1/2 at" (2) 

where o, the acceleration due to the force Ee 
acting on the electron between collisions is: 

a = — (3) 


Substituting equation (3) in (2) we obtain: 

2 m 
Since D= U t, we have from (4) 

1 =Ym- (5) 

now equation (5) applies to the electron dur- 
ing its motion between collisions, so that (t) 
is the average time between the impacts of 
all the electrons. From the kinetic theory 
of gases, 



where (h) is the mean free path of the elec- 
trons, and (v) their average velocity of vibra- 
tion over this path. Eliminating it\ from 
equations (5) and (6) we get: 
TT E e h 
2 m v 

Hence, when we substitute this value of I IT) 
in equation (1) we have: (See note at the 
end of this article.) 

I = N ** k (8) 

2 m v 

which is an expression for the intensity of the 
electronic current in any substance. If there 
are only a few free electrons (N) in a sub- 
stance, as in the case of glass, and other in- 
sulating materials, salts, as well as in electro- 
negative elements, there must, according to 
this theorv, be a relatively small current 



passing. On the other hand, if there are a 
great many free electrons in a substance, as 
in the case of the electropositive elements 
(metals), then there will be a relatively large 
current flowing. All this is in agreement with 
experimental observations. In this theory 
the atoms have played no part in the metallic 
conduction which, of course, is also in agree- 
ment with observation. If (E) is a com- 
plicated function, as in the case of high fre- 
quency phenomena, then equation (8) will 
be modified accordingly. In fact, (E) can 
represent a steady potential, a harmonic one, 
or any sort of a potential. It would be 
necessary, of course, to start with the fun- 
damental differential equations of motion for 
such an analysis of the problem. 

Before passing on to the thermal con- 
ductivity, it will be well to show that equation 
(8) is in agreement with Ohm's law. Before 
showing this, however, let us modify this 
equation. The kinetic theory of gases tells 
us that the average kinetic energy of a hydro- 
gen molecule at an absolute temperature 

(T) is a T= =- m v 2 , so that this is also, 

according to our theory, the average kinetic 
energy of the electron at an absolute tem- 
perature (T). (a) is a constant and equals 
1.5 XIO -16 approximately. Putting 2aT = 
w v- in equation (8), we have: 

L ~ 4:OCT 

and since the conductance <x is the current 
per unit electric force, 

<r=4^ (10) 

■i a 1 

Now, Ohm's law states that the conductance 
is independent of (£), which is in agreement 
with equation (10). It may be further 

stated that since resistance equal ( - ) equation 

(10) shows that the resistance varies directly 
as the absolute temperature. We may look 
upon resistance as being due to the collisions 
between the electrons and the atoms. One 
would also expect frequency to have an 
effect on the resistance because of the change 
in the number of collisions per cycle with 
increased frequencies. All these effects have 
been noted. 

If one part of a piece of metal is at a higher 
temperature than another, then the average 
kinetic energy of the electrons in the hotter 

regions will be greater than that of those in 
the colder portions. 

Considering that the electrons act like a 
gas, then it is clear that across any section 
separating the hot from the cold portions, 
a transfer of electrons will take place. Due 
to the greater kinetic energy of the electrons 
on the hot side of the section, there is a 
transfer of heat to the cold portion of the 
metal as the electrons pass from the former 
to the latter. If we assume that all the heat 
is carried in this manner, then it has been 
shown in works on the kinetic theory of gases 
that k, the thermal conductivity, is given bv: 
K=l/3 N v a h ' (11) 


Having obtained the expressions for the 
thermal and electrical conductance, let us 
see what relation exists between them. To 
do this we simply divide equation (11) by 
equation (10) thus, 

k , /„ »t i /N e-h v 

- = 1/3 Nv a h =r 

<j 4 a T 

k _ 4 a 2 T 
a~ 3e 2 

or at 7 = 300 deg. C, absolute. 

K = 4X(1.5) 2 X(10-' 6 )-X3X10 2 
a 3X(5X10- 10 ) 2 


4X10" 11 

approx. in C.G.S. electrostatic units. Equa- 
tion (12) shows us that the first theory of 
electronic conduction in metals leads us to the 
conclusion that the ratio of the thermal to the 
electrical conductivity should be the same for 
all metals, and should vary directly as the 
absolute temperature, being entirely inde- 
pendent of all metals. So that with good 
electrical conductivity goes good thermal 
conductivity. The following table shows how 
nicely these theoretical conclusions are veri- 
fied by actual experiments. 


(at 18° C. 

Temp. Coef. of — add 



0.043 X 10-" 


7.4 X10-" 

0.039 X 10"" 


7.6 XIO"" 

0.037 XIO"" 


8.0 XIO" 11 

0.036 X 10"" 


7.9 X10"" 

0.040 X 10"" 


8.3 XIO"" 

0.046 X 10"" 

We shall have occasion to point out some 
variations in this ratio in the next article. 



Since (k) is very nearly constant over a certain 
range of temperature, it follows from this 
theory that (<r) must decrease with increasing 
temperature; this agrees with experiment. 



It will be of interest to obtain an approxi- 
mation to the number of free electrons there 
must be in one cu. cm. of a metal according 
to this first theory. In the case of silver for 

example a = 



at deg. C. Now, we will 

not be far wrong if we assume the mean free 
path of the electron (h) to be about 10 -7 cm. 
Substituting the various values which have 
been given for the quantities involved in 
equation (10), we get N= 10 ;4 approximately, 
which is equivalent perhaps to five or six 
free electrons for each atom of silver. 

Perhaps it will be well at this point to 
refer to values obtained by other methods. 
J. J. Thomson concludes from work on the 
coefficient of absorption of radiation by a 
metal that the number of free electrons in 
silver appears to be not less than about 11 
for every atom of silver. Earlier than this 
Drude and Schuster in determining the ab- 
sorption of light by metals concluded that 
silver possessed about one, mercury about 
three and one-half, and antimony about 
seven and one-half, free electrons for each 
atom of metal. Work with Dulong and 
Petit's law, that the product of the atomic 
weight and specific heat is nearly the same for 
all metals and is constant, has led to the 
conclusion that the maximum number of 
free electrons is two per atom of metal. 
This conclusion is based on the assumption 
that all the heat energy be attributed to the 
free electrons associated w-ith the atoms. In 
the next article will be given another method 
by which we can estimate the number of free 
electrons in metals. Since the order of magni- 
tude for the number of free electrons in metals 
is the same when determined from such dif- 
ferent methods as here indicated, it would 
appear that there is some element of truth 
in this theory of free electrons. It will be 
well to note that equation (8) shows that 
the electric conductivity does not depend 
only on the number of free electrons in a 
metal. The value of (A 7 ) might be somewhat 
greater for poorer conducting metals than 
for good ones. The difference is accounted for 
by (h) and (v) being different for different 

It is very important to notice in connection 
with equation (10) that if (<r) is constant (as 
it is under given conditions), then (N h v) 
must be constant. Now, if the mass of the 
carrier of electricity was much greater than 
the electron, then (h v) would be less, and 
hence, (N) much larger than given by the 
above values. So that the number of carriers 
of electricity in this case would be much larger 
than the number of atoms of silver. We are 
forced to say, therefore, that these carriers 
cannot have masses comparable with those of 
the atoms, which likely take little part in the 
phenomenon of electric conduction. 

Perhaps it will be well to point out a very 
noticeable contradiction in the theory as 
developed here. The energy required to 
raise the temperature of an electron 1 deg. 
C. is a=1.5X10~ 16 ergs, about. So, if there 
are 10 24 free electrons in one cu. cm. of silver, 
then to raise the temperature 1 deg. C. of 
the electrons alone would require (1.5X 
10- 16 X10 24 ) = 1.5X10 8 ergs, or about 4 cal- 
ories. But to raise the temperature 1 deg. C. 
of one cu. cm. of silver requires only 0.6 
calories. Hence, the electrons alone use more 
energy, according to our theory, than do the 
electrons plus the atoms by actual experiment. 
(See note at the end of this article.) W.e 
must therefore modify the theory, somewhat, 
in order that it shall agree with the experi- 
mental fact. This change is made as follows: 


On account of this contradiction in the 
case of specific heats, it is quite certain that 
the electrons cannot be in thermal equilib- 
rium with its surrounding metallic molecules. 
In order to overcome this discrepancy J. J. 
Thomson has modified the above theory, and 
supposes that the electrons shoot from one 
atom directly into another one, and thus 
thermal equilibrium is not established between 
the atoms and electrons. This motion is in 
every possible direction, and hence, there is no 
resultant flow of current. (The positive and 
negative atoms form small electric doublets.) 
If, however, a potential is applied to the wire, 
then he considers the atomic doublets are 
polarized, much the same as some people 
consider the atoms of a permanent magnet 
are polarized. The negative sides of the 
doublets will be pointing in one direction, 
and the positive sides in the other. The 
result of this polarization is that more 
electrons move in one direction under the 
action of an electric force than in any other, 
so that a passage of current occurs. 



The final equations he obtained were: 
_ 2 e 2 d p nb 
a ~ 9 (a) T 
_ n b 2 p a 




3 b (a 2 ) T 




2 de 2 

Where (p) is the frequency with which a 
doublet liberates electrons, (n) is the number 
of polarized doublets per cu. cm., (b) is the 
distance between the charges in the doublet, 
and (d) is the distance between the centers of 
the adjacent doublets. In a metal (b) is 

nearly equal to (d) so that -j = 1 (approxi- 
mately). Hence, in this case the ratio of the 
conductivities on the new theory would be to 
that on the old in the proportion of 9 to 8 as 
given by equations (12) and (15). This 
theory as well as the old is in agreement with 
facts, but this theory tells us that this ratio 
is not an absolute constant on account of the 

factor ( -j J , which varies slightly for good con- 
ductors and more for bad ones. This is in 
agreement with fact. 


In this article have been mentioned the 
principle of the conservation of electricity 
and the three methods by which we con- 
sider it to flow. It has been pointed out how 
the electron theory applies to this important 
principle in that it explains the separation of 
the already existing charges. We have seen 
that this theory deals with convection cur- 
rents entirely, and that it explains fairly well 
both qualitatively and quantitatively the 
part now played by conduction currents, 
about which we know nothing. 

Although we have not developed an entirely 
satisfactory electron theory, as has been 
pointed out, yet, since we are learning more 
about electrons and the laws which they obey 
under all conditions, we are fairly certain that 
we are advancing toward a more complete 
understanding of the various pnenomena of 
electric conduction. 

Note : — A more rigorous treatment of 
this problem along the same line for a steady 
electric force (E) would lead to the result: 

1 = 2, 

N E e 2 h 


(See G. H. Sirens, "Electron Theory of 
Metallic Conduction," Phil. Mag. pp. 173- 
183, Jan. 1915.) This is also the same result 
that Lorentz obtained in his book on "The 
Theory of Electrons." H. A. Wilson (Phil. 
Mag. Nov. 1910) obtained still another 
expression for (7) ; it differs, however, from 
these only in the constant term. The above 
equation is about twice as great as indicated 
by equation (8). The above equation would 
lead to a value for (A r ) about half as great 
as does equation (8). Even this, however, 
would not yield values for specific heats 
that were in agreement with observations. 

In connection with the question of specific 
heats one may refer to Lindemann, "Theory 
of Metallic State" Phil. Mag., p. 129, Jan. 
1915. In this article the author states. 
"The hypothesis put forward in this paper is, 
that far from forming a sort of perfect gas the 
electrons in a metal may be looked upon as a 
perfect solid." Statements similar to this 
indicate that our conception of the electron 
itself has not changed so much as has our 
conception of its intimate association with 




By Lewis A. Mason- 
Assistant Designing Engineer in Office of the Engineer of Maintenance of the Panama Canal 

Eight articles describing the devices controlling the lock machinery of the Panama Canal appeared in 
the January, 1914, number of the General Electric Review; three articles describing the generation and 
distribution of electric power for the Canal Zone were contained in the July, 1914, issue; and the towing system 
and locomotives were described in an article in the February, 1915, issue. The following pages present an 
interesting description of the mechanical, hydraulic, and electrical features of the lock entrance caisson which 
is to be used at the locks to hold back the water and to pump out the chambers when repairs are to be made to 
apparatus that is normally submerged. — Editor. 

In connection with the various equipment 
required for the maintenance of the Panama 
Canal Locks, the Union Iron Works Company, 
of San Francisco, has recently completed 
a huge floating gate or caisson which will 
be used for closing the entrance to any one of 
the lock chambers of the Panama Canal 
when it is desired to paint or make repairs 
to any one of the mitering lock gates and for 
similar use in the Balboa dry dock. It also 
can be used for unwatering any one of the 
lock chambers, for the purpose of making an 
inspection of the culvert, rising stem gates, 
or cylindrical valves. 

The clear width of the lock chambers is 
110 feet. Beyond the line of the emergency 
dams, the approach is widened by an offset 
of three feet on both sides. The shoulders 
so formed, with the connecting horizontal 
sill across the bottom of the chamber, afford 
a frame or seat into which the caisson is 
fitted to dam off the interior of the lock 

This is accomplished by floating the caisson 
from its mooring position by means of a tug 
boat, or other motive-power water craft, to 
the particular lock chamber entrance which 
is to be dammed. After being placed in its 
recess across the lock entrance, water will 
be let into the lower compartments, thereby 
causing it to sink until properly seated. 
When this is completed, an electric power 
cable will be connected from the main power 
cables, provided within the lock walls, to 
a terminal box located on the top deck and 
at the end of the caisson. This point is 
electrically connected through the switch- 
board within the caisson to the various 
motors that operate the pumps. The pumps 
will then un water the lock chamber, and the 
water pressure on the outer side of the caisson 
will force it securely against its seat in the 

When it is desired to remove the caisson, 
the lock chamber will first be filled with 
water by opening the culverts within the 

lock walls. This will balance the water 
pressure on both sides of the caisson, at 
which time the water within it will be pumped 
out, thereby causing it to float and allow 
it to be towed awav. 

Fig. 1. 

An End View of the Caisson Taken but a Short 
Time Before it was Launched 

The caisson is designed for use at all of 
the lock entrances, and has a light-draft of 
32 feet to permit its being handled convcn- 









iently through the various locks. The top 
of the sill at the Pacific end of the Mira- 
flores locks is 50 feet below mean sea-level. 
and with the tidal fluctuation which raises 
the level of the water as high as 11 feet 
above mean tide this requires that the caisson 
be sunk to a draft of 61 feet when used at 
high tide. Provision for a proper freeboard 
requires an aggregate depth of the structure 
of 66 feet. The achievement of statical 
stability at the various depths of immersion 
without undue bulkiness or excessive weight 
in the different drafts makes the caisson of 
especial interest. 

In form, the bottom of the hull is convex, 
the ends pointed, and the sides sloped inward 
from the maximum width of 36 feet, at about 
one-third the way up from the keel, to a 
breadth one-half as great at the top deck. 

framing built intercostally and extending 
from the keel to the top deck transmits the 
panel loading to the various horizontal 
decks and breasthooks. The essential features 
of the structure are the transverse and 
longitudinal framing, with bulkheads; the 
horizontal plate decks, girders and stringers; 
the girders at the vertical ends and along 
the keel; the end breasthooks; and the sheath- 
ing plates to cover the skeleton for forming 
the hull proper. These elements are made 
from open-hearth structural steel. 

The transverse framing system consists 
of nine cross-frames, spaced 12 feet apart 
from the middle of the caisson and extending 
to its entire height, and the intermediate 
frames, spaced two feet apart between the 
main cross-frames. All are built intercostally 
between the five horizontal decks. 

Fig. 6. Plan View of the 37 Ft. Deck Showing the Location of the Electrical Apparatus 

A typical transverse cross-section of the 
caisson resembles in outline the vertical 
section through a pear-shaped, carbon-fil- 
ament electric globe. The horizontal length- 
wise sections vary with the inward slope of 
the sides; in general, they resemble those of 
the ordinary vessel of commerce, and may 
be described as flattened ellipses, blunt at 
the ends in order that they may connect to 
the vertical end-girders, or stems. The 
maximum length of the caisson from vertical 
end to vertical end is 112 ft. 6 in. The 
extreme length is 113 ft. 10 in. This includes 
the timber cushions. 

It is ^desired that the side walls of the 
locks shall carry practically all the static 
load from the caisson when it is supporting 
the water pressure. Accordingly, there are 
a number of horizontal decks and breasthooks, 
or short decks, between the main decks at 
'he ends which carry the hydrostatic load 
to the vertical ends A system of vertical 

The last cross-frame at each end is made 
water-tight, by the same principle as is used 
in merchant ships, in order to form peak 
trimming tanks for maintaining a level 
keel when placing the caisson in its recess 
across any one of the lock chambers. The 
seven other cross-frames serve as swash 
bulkheads for controlling the water within it. 

The five horizontal decks are located at 
the respective following distances above the 
centerline of the keel plate: 16 ft., 25 ft., 37 ft., 
49 ft., and 65 ft. The 16 ft. and 25 ft. decks 
are entirely plated over with the exception 
of openings left to allow for the removal of 
pumps, valves, etc. 

The 37 ft. deck is entirely plated over 
and is made absolutely water-tight. It 
has water-tight manholes for gaining access 
to the various compartments below and 
water-tight hatches for the removal of the 
pumps or valves in case it is necessary to make 
repairs, etc., to them. This deck is made of 



sufficient strength to withstand a hydrostatic 
head of 25 ft. Upon it is placed the various 
motors for operating the pumps, the switch- 
board, the water gauges, the chain lockers, 
etc. The horizontal deck 49 ft. above the 
center line of the keel is of the open-truss 
construction, and has diagonal bracing for 
the central two-thirds of its length and 
plating covering for the ends. The top deck, 
65 ft. above the center line of the keel, is 
plated over from end to end and has openings 
for manholes, skylights, deck cranes, com- 
panionways, ballast compartment vent pipes, 
and scuppers. 

The breasthooks, or short decks, of which 
there are six in number, serve to transmit 
part of the loading from the horizontal 
decks to the vertical end-girders. In addition 
to the decks and breasthooks, there are 
located equidistant between the keel ■ and 
the 16 ft. horizontal deck two lines of inter- 
costals extending longitudinally and securely 
riveted to the transverse frames and to the 

For transmitting the end reactions from 
the horizontal decks and breasthooks to the 
vertical end-girders, or stems, steel castings 
are provided and made to fit very closely 
between the horizontal decks and breast- 
hooks and the vertical ends, to which they 
are securely riveted. 

The skeleton or framing is entirely sheathed 
over with steel plating worked in in-and-out 
strakes, running longitudinally over the 
transverse frames, making lap seams and 
butt joints which have double splice plates. 
Around all of the openings in the plate 
decks, and in the sheathing, doubling or 
reinforced plates are fitted. To protect the 
sheathing when maneuvering the caisson 
near the lock walls, fenders are provided 
on the exterior of the sheathing along the 
25 ft. and 49 ft. levels, and vertical fenders 
are placed between the horizontal ones, at 
seven of the amidship cross-frames. The 
fenders are built of bent plates, securely 
riveted and calked to the sheathing plates; 
the space between is filled with "Petro- 
lastic" cement — a by-product of crude oil. 
Its specific gravity is 1.02; its expansion 
at a temperature of 110 deg. is 0.0018, and its 
melting point lies between 150 and 200 deg. F. 

Because of the long towing distance from 
the place where the caisson was built two 
large towing rings are provided and are 
securelv fastened to the sheathing and to 
the 43" ft. breasthook at both ends and on 
each side of the caisson. As a means for 

towing the caisson from its mooring position 
to any one of the lock sites, there are three 
towing rings provided which are securely 
riveted to the sheathing along the level of 
the 37 ft. horizontal deck on both sides of 
the caisson. 

Along the exterior of the keel and the 
vertical ends, steel castings (the cross-sections 
of which are channel-shaped) are provided 
and are securely riveted to the keel, vertical 
ends, and sheathing. Into these there are 
neatly fitted and bolted British Guiana 
greenheart and Australian ironbark timber 
cushions. There is also a cushion fitted 
along the sides of the keel and along the sides 
of the vertical ends, which are also made of 
the timbers mentioned. These cushions 
come into contact with the caisson's seat 
provided in the lock chambers and form a 
water-tight seal. 

Miscellaneous Fittings 

Through a water-tight companionway on 
the top deck a stairway leads down to the 
37 ft. operating deck. Ladders from this 
deck are provided for gaining access to the 
various lower compartments. Ladders are 
provided in the end peak trimming tanks, 
extending from manholes in the top deck 
to the 16 ft. horizontal deck, or bottom of 
the trimming tank. For getting aboard the 
caisson a ladder is provided on each side 
and is attached to the sheathing. It extends 
from the level of the 32 ft. water-line to the 
top deck. 

There are three portable cranes located 
on the top deck, one at each end of the 
caisson and one in the center. The two end 
cranes are similar in construction, and are 
capable of raising or lowering a load of 3000 
lb. at a radius of 14 feet by two man-power. 
These cranes are used for lifting various 
loads onto the caisson from the lock walls, 
as well as for handling electric power cables. 
The middle crane is heavier in construction 
than the end cranes and is capable of raising 
or lowering a load of 3000 lb. at a radius of 
25 feet by two man-power. This crane will 
handle the pontoon (stowed on the top deck ) 
when it is desired to make the pump suction 
extension attachments, and is capable of 
lifting the top sections of either one of the 
two skylights. Hand-operated deck capstans 
are provided and placed at each end of the 
caisson on the top deck. The capstans are 
installed for the purpose of warping the 
caisson into its recess. Each is capable of 
withstanding a pull of 10,000 pounds. 



Two ventilators, each 16 inches in diameter, 
with hoods and turning mechanism of the 
standard navy type, are placed on the top 
deck for ventilating the operating room. 
Both of these ventilators extend from the 
top deck to a short distance below the 49 ft. 
horizontal deck. At the end of one is fitted 
and connected an electric-driven multivane 
exhauscer to supply a means for assisting 
the air to escape from the various water- 
ballast compartments when they are being 
filled. There are eight 6-inch diameter air 
vents, extending from the various ballast 
compartments to the top deck, and one air 
vent in each of the end peak trimming tanks 
placed in the top deck. Two skylights 8 ft. 
by 16 ft. in size are fitted in the top deck, 
symmetrical about the axis of the caisson. 
The tops are made in two parts, for easy 
removal. In each top section there are 
openings fitted with water-tight covers, 
which can be opened or closed by means of 
a raising apparatus located and secured 
under the top deck and operated by means 
of a handwheel from the operating deck. 

To increase the draft of the caisson to a 
depth sufficient to insure its stability at 
light draft, without water in the ballast 
compartments, approximately 800 tons of 
permanent ballast, composed of iron punch- 
ings, etc., and concrete, is placed in the 

An anchor chain, made of material 1^ 
inches in diameter, is provided at each end 
for mooring the caisson to floating buoys 
in the fresh water lakes when it is not in 
service. The anchor chains are raised or 
lowered by means of the hand-operated 
winches, located at each end on the top deck. 

Pumping System 

The main pumping system consists of 
four vertical-shaft, bottom-suction type cen- 
trifugal pumps which, with their individual 
driving motors, constitute four units. Each 
unit is designed for an average capacity of 
13,000 g.p.m. against a maximum head of 
70 ft., this capacity being the average to 
prevail between heads varying from zero 
to the maximum (7(1 ft.). The suction 
opening of each pump is 22 inches in diameter 
and the discharge 20 inches. 

From the illustration of the outline drawing 
of the complete pumping unit, it will be seen 
that the intermediate shaft connecting the 
pump to the driving motor is supported by 
an intermediate guide bearing. The drawing 
at the thrust bearing, which 

carries the load of the revolving element, is 
located at the motor deck and is contained 
in a base-plate which, in turn, acts as a support 
for the motor itself. The thrust bearing is 
of the ball-bearing type, and is made self- 
oiling by means of an oil pump which takes 
its supply from a revolving pin located 
beneath the bearing and which returns the 
oil to a reservoir that surrounds the ball 
bearing. The intermediate guide bearing 
and the pump bearing are water lubricated. 
The pump casing, together with the impeller, 
is made of cast-iron and' is bronze lined at the 
points where the impeller comes in contact 
with the casing, also where the shaft passes 
through the bearing and the stuffing box. 

The pumping plant is employed for a 
double purpose: first, for emptying the 
water ballast from the caisson when it is to be 
removed from its position against the sill 
and, second, for unwatering all lock chambers 
except those which can be emptied by gravity. 
(The only chambers in the Panama Canal 
that can be emptied by gravity are the upper 
lock chambers at Gatun; the elevation of 
the floor there is 13% ft. above sea level.) 

The capacity of the pumping system is 
designed so that it will pump out, in not 
more than 25 hours, all of the water in the 
upper and lower chambers of one flight of the 
Miraflores locks between mean sea-level 
(elevation zero) and the top of the sill of 
the lower chamber ( — .30 ft); the tidal level 
to be at elevation zero when the pumping 
is begun and the tide to be rising. The 
total quantitv to be pumped is estimated 
at 10,285,000" cubic feet. Of this quantity 
5 IS, 000 cubic feet is allowed for leakage 
through the various cylindrical and rising 
stem gate valves in the lock culverts, as well 
as allowances for leakage around the sills of 
the mitering lock gates arid the caisson sill. 
The pumps will pump out, when operating 
at any stage of the tide, the water on the 
floors of the lower lock, from the top of the 
sill (-50 ft), to 2 ft. below it. To do this 
22-inch suction pipes are attached to the 
auxiliary suction inlets of the caisson, and 
these extend to and into the nearest lateral 
culvert in the lock chamber. When not in 
service, the suction extension pipes, of which 
there are four in number, are stowed in 
cradles provided for the purpose on the 49 ft. 
horizontal deck. They are handled by the 
large deck crane located in the center line 
of the top deck. 

An electric-driven horizontal centrifugal 
pump, with a 3V 2 in. diameter suction and 



a 3 in. discharge, is located on the operating 
deck. It has pipe connections leading from 
the suction to a manifold and from the 
discharge to another manifold. From these 
manifolds piping is connected to the end 
peak trimming tanks, to the deck scuppers, 
to the sea, and to a mud-slushing device. The 
mud-slushing device is intended to remove 
mud from the caisson sill in an endeavor to 
prevent it from adhering to its seat when in 
the act of rising. The pumping equipment 
was manufactured by Henry R. Worthington, 
Harrison, N. J. 

Electrical Equipment 

The main pumps are driven by 200-h.p. 
vertical induction motors, wound for 25 
cycles, 2200 volts, three-phase, and which 
have a speed of 750 r.p.m. The motors for 
operating the ventilating fan and the 3-inch, 
or auxiliary pump, are induction motors of 
the horizontal type, and arc wound for 25 
cycles, 220 volts, three-phase. For lighting 
purposes, 110 volts are used. All of the elec- 
tric motors (with the exception of the one for 
driving the multivane exhauster, and their 
controlling switchboard are located on the 
operating deck, 37 ft. above the base line. 
All of the valves in the pumping system are 
operated from this same deck. 

The switchboard consists of five panels, 
and, from right to left, facing the front of 
the board is arranged as follows: One, 
three-phase, three-wire, incoming line panel; 
two, three-phase, three-wire, double-circuit 
motor panels; one, three-phase, three-wire, 
motor feeder and lighting, transformer panel ; 
and one, single-phase, two-wire, ten-circuit 
lighting panel. Grille work having hinged 
doors provided with locks enclose the ends of 
the board and prevent access to its rear 
except by those authorized persons who are 
furnished with a key. From the top of the 
panels, and extending upward for some 
distance above the busbars, grille work is 
also provided. 

The arrangement of the switchboard ap- 
paratus, including bus and connection bars, 
is especially compact and is supported 
in a most substantial manner. This can 
easily be seen from the back view of the 
installation. The bus and connection bars 
are of three-quarter inch solid copper rod, 
the connection bars being soldered into 
terminals which are fastened to the busbars. 
The busbars are supported to the pipe frame- 
work by special bus supports designed for 
use in connection with this and other Panama 

Canal switchboard installations by the Gen- 
eral Electric Company of Schenectady, N. Y., 
which also supplied the electrical equipment 
for the caisson. The framework itself is of 
standard type, but was specially galvanized 
and painted to enable it to withstand the 




Z4Ft 7,1/n 

Fig. 7. 

Elevation of the Pump Motor, its ] 
Bearings and the Pump 


particularly severe climatic conditions pre- 
vailing on the Isthmus. 

Another feature of interest is the method 
employed for disconnecting the oil switches 
from the circuit, when it is desired to remove 



Fig. 8. Rear View of the Operating Switchboard 

the oil cans or otherwise to do work about 
the back of the board. By means of handles 
located below the oil-switch operating handles, 
a switch can be placed in or disconnected 
from its circuit at will whenever the oil switch 
contacts are open, but at no other time. 

The pipe framework supports 
vertical metal guides, which carry 
the oil-switch operating mecha- 
nism, and a slate base which 
forms a portion of the switch- 
board panel. By means of a lever 
and toggle mechanism, the oil 
switch, the slate base, and the 
other parts carried on the guides 
may be raised or lowered. 

Above the oil switch and mounted 
on the pipe framework is a station- 
ary base which carries the discon- 
necting studs of the oil switches. 
The current leads are connected 
to the tops of these studs; and at 
the bottom of each stud is a flared 
contact which engages with a wedge- 
shaped contact on the upper end 
of the oil switch stud, and thus 
places the switch in circuit. Moulded 
insulating shields surround (except 
at the bottom) each disconnect- 
ing contact and extend sufficiently 
below the contact fingers to insu- 
late the fingers and prevent acci- 

dental contact, whether the oil 
switch is or is not disconnected from 
the circuit. 

The oil switch can not be con- 
nected to or disconnected from the 
circuit except when it is in the 
open position, which guards against 
the circuit being closed or opened 
by the disconnecting contacts. This 
feature is made possible by an inter- 
lock that prevents the oil-switch 
lifting and lowering handle from 
being operated unless the oil-switch 
operating handle is in the open 
position. This oil switch arrange- 
ment is the development of a patent 
by Mr. E. Schildhauer, Electrical and 
Mechanical Engineer of the I.C.C. 

The electric current is supplied 
to the motors in the caisson from 
the main power cables installed 
within the lock walls. The motors, 
therefore, cannot be operated until 
the caisson is seated in one of the 
recesses provided for it in the locks, 
or when at its mooring position in Gatun 
Lake or Miraflores Lake. 

The purpose for having a power connection 
at its mooring position is to permit the opera- 
tion of any one of the pumps for examination 
and inspection. 

Fig. 9. Front View of the Operating Switchboard 



Part VI (Nos. 32 to 35 inc.) 

By E. C. Parham 
Construction- Department, General Electric Company 


Motor overloads are sometimes due to 
rather unexpected causes. Electrical inspec- 
tors are prone to exhaust the possibilities 
of electrical diagnosis, in times of trouble, 
before looking for mechanical irregularities 
that would account for unusual actions in 
electrical apparatus. 

An inspector was called to find out why the 
starting resistor of a certain three-phase 
induction motor driving a monorail crane 
would get white hot whenever an effort was 
made to operate the crane. 

The crane had just been installed and it 
was, of course, to be expected that the start- 
ing resistors would heat somewhat more than 
normally, because the crane action is stiff 
and the numerous bearings have not found a 
seating. In this particular case, however, 
the extent of the heating, considering the 
promptness with which the crane would 
start, suggested a condition more serious 
than initial stiffness. 

In order to make an electrical test of the 
crane wiring, it was deemed advisable to 
insulate the crane from its source of power by 
inserting thin sheets of insulating fiber 
between the three overhead contact rails and 
the corresponding contact-shoes which are 
pressed against the rails by means of springs. 
To pry the shoes down from the rails required 
two men with two four-foot jimmies. What 
springiness there was to the contact-shoe 
action was due to upward pressure springing 
the 2 in. by ?g in. T-sections of which the 
contact rails were made. When the shoes 
happened to be directly under an insulator, 
there was no spring action at all. This con- 
dition of affairs was due either to the shoe- 
stands being too high or to the contact 
rails being too low; whatever was the 
cause, the crane was being continuously 
subjected to the retarding action of a strong 
track brake. 

Considering the fact that some high-speed, 
third-rail, electric railway cars use shoe 
pressures that do not exceed 30 pounds per 

shoe, the load to which the two 3-h.p. crane 
motors were being subjected may be appre- 


It would seem that the harder the service 
and the more exacting the local conditions 
to which electrical appliances are subjected, 
the greater the equipment is neglected. 
Where fastenings are most likely to be shaken 
loose by unavoidable vibrations, inspections 
for loose parts seem to be most lax. Really, 
the opposite should be true. These impres- 
sions are justified by the frequency with 
which solenoid-braked foundry crane-hoist 
motors giving slight troubles are permitted 
to become serious troubles simply for want of 
the prompt detection that would result from 
regular inspection for loose parts. 

An operator once complained that his 
crane-hoist motor had "stuck with a pot of 
metal in the air." Inspection showed that 
if the motor had not stuck, it probably would 
have been wrecked. The normal adjustment 
of the air-gap of the solenoid brake was from 
% in. to 1 in. The gap had been allowed to 
become 3 in. As a result, the hammering at 
the brake end of the motor, when the brake 
operated, loosened every bolt on that end of 
the motor. The end-shield bolts had worked 
entirely out, notwithstanding the fact that 
they had been secured by lockwashers, which 
had let the rotor down onto the stator. 
Fortunately the revolving magnetism did 
not provide sufficient torque to turn the rotor 
in this locked position; otherwise both rotor 
and stator probably would have been ir- 
reparably damaged. 

The adjusting mechanism of a solenoid 
brake is not complicated, and the attention 
that it requires is not the attention of an 
electrician but that of a crane-man. Nearly 
every concern of sufficient size to use a power 
crane has at least one mechanic qualified to 
detect when a piece of apparatus is shaking 
itself to pieces. Most operators, however, 
seem to dissociate electrical apparatus from 



every-day common sense relief measures 
which are always worth at least a fair trial. 


If the fuses used are not too large, the first 
indication that an induction motor's rotor is 
rubbing its stator, may be the melting of the 
fuses because of the increased load incident 
to the additional mechanical friction. 

A certain 5-horse power, three-phase induc- 
tion motor sometimes would start upon 
applying the power and sometimes it would 
not start; but the stator would give the 
characteristic single-phase hum and the 
rotor would oscillate through a very small 
arc suggesting that it might start in either 
direction. All of the windings and wiring 
proved by test to be free from open-circuits, 
grounds, and wrong connections, and the air 
gap (which was normally 0.015 in.), freely 
admitted a 0.01 in. "feeler" all around the 
rotor and at both ends. These tests were 
made with the motor disconnected by the 
removal of its pinion. Upon reinstalling the 
pinion and wedging the gear, so that it could 
not move the load, and then applying the 
power and observing the rotor closely, it 
could be seen that upon each application of 
the power the rotor would move upward in a 
direction corresponding to the direction of 
the force applied to the gear by the pinion. 
On most trials, with the wedge withdrawn, 
the movement was insufficient to cause the 
rotor to touch the stator and on such occasions 
the rotor would start, and after it was in 
motion no irregularity could be observed. 
Now and then, however, the rotor would rise 
far enough to stick and it would start only 
upon advancing the controller to farther 
notches. The lifting of the rotor indicated 
no bearing wear, but by removing the rotor 
and linings a test of their fit showed a slight 
wear just at the place that would account 
for the symptoms noted. This slight wear, 
probably in conjunction with a slight eccen- 
tricity in the rotor core, accounted for the 
fact that sometimes the motor would start 
and sometimes it would not. The rotor sur- 
face had so much oil on it that a slightly 
rubbed place could not have shown very 
plainly, but such a place undoubtedly existed, 
and when this place was in line with the bear- 
ing wear at the time of applying power the 
rotor would stick. 

The lesson to be drawn from this experience 
is that the air gap of a motor at rest may be 
thoroughly correct as far as a feeler will 

indicate, but, if the bearing wear is in the 
upper part of the linings and the direction of 
rotation is such as to force the rotor upward 
at starting, the rotor may strike the stator. 


In some classes of foundry work that is 
handled by electric cranes a smoothly grad- 
uated motor acceleration is essential; espe- 
cially is this the case during the period of 
separating the cope from the flask, for then 
an impulse may shake down sand and destroy 
the mould. The binding between the cope 
and the flask complicates matters, because it 
introduces a condition where a comparatively 
strong pull must be immediately followed by 
an easing off, which is not always to be 
obtained satisfactorily. Where a variety of 
work is to handled smoothly, it is necessary 
to use a controller that has many notches. 
Where the weights to be handled are limited 
to a few standards, the resistance graduations 
can be refined to suit the weights involved. 
In either case, wide fluctuations in the 
supply voltage make it difficult to get equal 
degrees of smoothness for all weights and 

An electric crane, the hoist control of which 
had been entirely satisfactory for a long time 
but which had been getting more and more 
jerky during a period covering about two 
years, finally became impracticable and an 
inspector was called in. The crane evidenced 
good care, as far as the crane man was con- 
cerned, and the controller fingers and con- 
tacts were in excellent condition. Being 
convinced that neglect was not the cause of 
the impulsive acceleration, the resistor was 
next investigated. Tests with a voltmeter 
showed widely varying voltage drops per 
section of the resistor, and one of the sections 
caused no drop at all. The resistor was 
housed in a perforated box that served also 
as a seat for the crane man. Upon removing 
the box cover and sides to inspect for bad 
connections and for broken and short-cir- 
cuited grids, the causes of the^-impulses be- 
came evident — they were two files, a screw- 
driver without a handle, a cold chisel, an 
oblong roll of copper wire, two carbon 
brushes, and seven perfectly good cartridge 
fuses. Without these the resistor was all 
right, as was demonstrated by a trial. To 
prevent a repetition of such a condition in 
the future, a piece of one-quarter-inch mesh 
wire netting was fastened to the under side 
of the resistor cover before replacing it. 


By J. H. Torrens 

The author gives a brief account of the conditions existing in Guatemala where native labor still 
performs many operations now done by machinery in more developed countries. He then proceeds to give 
a description of the hydro-electric plant of the Finca Ona Plantation which is the largest in Guatemala. The 
process of preparing coffee for the market and the different operations carried out by the aid of electric 
motors are described. — Editor. 

Guatemala is the largest, the most thickly 
populated, and probably the furthest devel- 
oped of any of the Central American Repub- 
lics. Although more than the usual quota 
of tropical products are raised there, the 
industry of coffee growing is the one of 
paramount importance. 

Before entering into a description of 
the hydro-electric installation that will be 
considered, a few remarks concerning the 
geography of Guatemala and the conditions 
existing there will be of educational service 

while they own as much as 85 per cent of the 
coffee estates. 

The primitiveness of the transportation 
facilities will be easily comprehended when 
it is considered that most of the freight 
is carried on the backs of Indian porters. 
These bearers will jog along easily at a five- 
mile-an-hour pace with a pack of 150 lb. 
and are ably capable of managing packs 
weighing as much as 200 lb. 

The work on the coffee plantations is 
carried on by native Indians, and these 

Fig. 1. View Showing the Old Rope Drive Transmission 

, Fig. 2. Hydro-electric Power Plant at Finca Ona 

in presenting a conception of the great and 
practically unentered field of coffee plan- 
tation electrification. 

American capital is responsible for the 
railways which connect Puerto Barrios of 
the east with the capital, Guatemala City, 
and witli San Jose and Ocas on the Pacific 
coast. Except for the omission of a few- 
miles of track between Coatepeque and 
Pajapita (as a matter of fact this gap is now 
nearly bridged) the Pan-American Railroad 
makes it possible to travel from New York 
City to the capital of Guatemala, which is 
at an altitude of 5000 feet. 

The commerce of the country is largely 
under the control of European countries; and 
it is said their investments in coffee plan- 
tations alone amount to about $60,000,000, 

receive about eleven cents a day for their 

Located in the western part of the country, 
among the foothills of the Sierra Madre 
Range at altitudes of from 2000 to 4000 feet, 
are some of the world's best coffee plan- 
tations. The climate in that section is par- 
ticularly suited to coffee growing. The Finca 
Ona plantation, which is situated in that 
vicinity, is one of the largest in Guatemala, its 
annual production being about 1,000,000 lb. 

The owners of this estate recently decided to 
adopt electric drive for the various machines 
used in preparing coffee for the market, and, 
as the conditions on this plantation may be 
considered to be typical of others throughout 
the coffee growing districts a description of its 
electrification should prove interesting to 



Fig. 3. View of a Transmission Pole and a Group of 
Coffee Trees 

Fig. 5. Revolving Drum used for Drying the 
Coffee Berries 


Fig. 4. Electric Driven Retrilla or Coffee Huller 

Fig. 6. Method Employed in Mounting the Motors 


the coffee planter, to the manufacturer of 
coffee-milling apparatus, and to the electrical 
engineer as well. 

The original source of power for the planta- 
tion was a Pelton waterwheel, for which it 
was necessary to bring water a distance of 
nine miles in a ditch. The difficulties which 
arose in the wet season from this method of 
water supply can be easily imagined. 

The power generated by the waterwheel 
was then transmitted 600 ft. by a rope drive. 
Two steam engines, one of 25 h.p. and the 
other 60 h.p., supplemented the waterwheel. 
The extremely difficult conditions of trans- 
portation from the railroad 30 miles away 
rendered the cost of imported fuel almost 
prohibitive, while the practice that had 
cleared all timber from valuable coffee 
lands made wood for fuel quite scarce. 
Consequently, power generated by steam was 
very expensive. 

A site at a convenient water-fall, which 
was about a mile from the factory, was 
chosen for the location of the electric 
generating station. From there the power 
is transmitted to the various motors and 

The hydraulic development was designed 
for about 500 cu. ft. of water per minute at 
an effective head of 270 ft., through 780 ft. of 
18 in. pipe to two Pelton waterwheels mounted 
on same shaft and rated at 230 h.p., 450 r.p.m. 
A Pelton self-contained, oil-pressure governor 
regulates the speed by the deflecting-hood 

The electrical apparatus in the power- 
house consists primarily of one revolving- 
field, 16-pole, 150-kv-a., 450-r.p.m., 2300-volt 
alternator direct-coupled to the waterwheel. 
The exciter is mounted on the same shaft. 
Frequent earthquakes make it imperative to 
mount the machines in a very substantial 
manner on heavy stone and cement founda- 
tions. All wiring is carried in conduit to a 
blue Vermont marble switchboard mounted 
on standard pipe framework. There are two 
feeder panels, one supplying 120 kw. at 2300 
volts to the main factor}* over a transmission 
line about a mile long, the other supplying 
a branch factory about one-half mile away 
with 15 kw. at 2300 volts. Both lines are 
thoroughly protected from lightning, first, 
by a well-grounded barbed-wire running 
from pole-top to pole-top throughout the 
entire distance, and, second, at both ends 
by the latest type of aluminum-cell elec- 
trolytic lightning arresters. For transmis- 
sion poles, 35-lb. iron rails 30 ft. long were 

used. These were "footed" five feet in the 
ground in concrete. 

At the factories the power is transformed 
to 220 volts for both motors and lights. All 
motors are of the three-phase squirrel-cage 


Front View of the Distribution Switchboard 

type, complete with starting compensators, 
and designed for a no-load speed of 600 r.p.m. 

To follow the coffee through the various 
processes in its preparation may be interesting 
as well as illustrative of the application of 
the motor drive. A 20 h.p. induction motor 
drives a battery of peeling machines to which 
the red coffee berries are fed as they come 
from the trees. These machines remove the 
tough outside skin and separate the two 
berries, or halves, which then go to the 
fermentation tanks where they remain in 
water for about 60 hours. 

At the end of this time, the thin membrane- 
like skin about the berries begins to loosen. 
All the good berries (those which float are 
not good) are taken out and placed in the 
sun on the "patio" where they are dried (by 
constant turning) to the extent that they 
cease to adhere to each other. This process 
requires several hours. 

After the berries are superficially dried 
in this manner they are placed in a large 
sheet-iron cylinder called "the drier." 

Those at Ona are about 6 ft. in diameter 
and 12 ft. long and through them steam- 
heated air at about 60 deg. C. is forced by 
a blower. In this cylinder the coffee is 
continuously revolved at 15 r.p.m. for 24 
hours; this operation is a particular one and 



requires close attention. It leaves these 
drums perfectly dry. and is then raised to 
the top of an ingenious machine which 
removes and carries away the hulls, polishes, 
and cleans the berries. From here the coffee 
is carried to the classifying and separating 
machines, the better grades being further 
sorted by hand labor. Each drier, huller, 
and separator has its own motor, the respec- 
tive capacities being 10, 20, and 10 horse 
power. All the starters are conveniently 
grouped near the main switchboard, where 
each motor has also an ammeter. 

Among other useful motor applications on 
the plantation might be mentioned a 1.5-h.p. 
motor, direct-geared to a Goulds' triplex 
pump which raises water 180 ft. for general 
use. This motor is operated also from the 
main switchboard. It is only necessary to 
visit the machine from time to time to see 
that the bearings are properly lubricated. 
A 2 h.p. motor which drives an ice machine is 
another valuable adjunct. 

All the electrical apparatus is three-phase. 
(iO-cycle, and was manufactured in America. 
Because of the crude and primitive methods 
by which the apparatus would have to be 
transported, it was necessary that the design 
employed be one that would permit of the 
apparatus being conveniently dismantled 
and, in addition, limit the maximum weight 
of a single piece to 1500 lb. Transportation 
part of the way was carried on by teams of 
bulls hauling a crude cart which was arranged 
so that when its two rear wheels were removed 
the rear half rested on a pair of sled runners. 
These latter were used to secure a braking 
action on the steep down grades. In other 
places the apparatus was carried on the backs 
of Indians. Thirty men carried a 40-kw. 
transformer weighing 1500 lb. in this 

The best construction possible was used 
throughout, and the work of installing all 
the apparatus was carried to completion in 
about three months. 

The electrification holds a rather unique 
position as it is one of the first installations 
of American apparatus in that country 
where European apparatus and interests 
predominate. It is also the first electrifica- 
tion of a coffee plantation of importance and 
is consequently being closely watched. 

Undoubtedly the continuance of the un- 
usually successful operation already enjoyed 
by this plantation, since its electrification, will 
induce other coffee growers to duplicate the 
change on their plantations. 

OF THE A. I. E. E. 

Third Mid- Winter Convention 

On Wednesday, Thursday and Fridav. 
February 17th, ISth and 19th, the third 
New York Mid-Winter Convention of the 
American Institute of Electrical Engineers 
was held at Institute headquarters, Engineer- 
ing Societies Building, 29-33 West 39th 
Street, New York. 

A very attractive program was prepared, 
including a number of pleasure events as 
well as an interesting selection of papers. 
The following papers were presented : 

The Characteristics of Electric Motors In- 
volved in their Operation, by D. B. 

Effect of Moisture in the Earth on Tempera- 
ture of Underground Cables, by L. E. 

Oil Circuit Breakers, by K. C. Randall. 

Comparison of Calculated and Measured 
Corona Loss Curves, by F. W. Peek, Jr. 

A 100,000-Yolt Portable Substation, by 
Charles I. Burkholder and Nicholas 

Distortion of Alternating-current Wave Form 
Caused by Cyclic Variation in Resistance, 
by Frederick Bedell and E. C. Mayer. 

Dimmers for Tungsten Lamps, bv Alfred E. 

Searchlights, by C. S. McDowell. 

Electrical Precipitation — Theory of the Re- 
moval of Suspended Matter from Fluids, 
by W. W. Strong. 

Theoretical and Experimental Considera- 
tions of Electrical Precipitation, by A. F. 

Practical Applications of Electrical Precipi- 
tation, by Linn Bradley. 

Electrical Porcelain, by E. E. F. Creighton. 

Institute Library 

The American Institute of Electrical Engi- 
neers, the American Society of Mechanical 
Engineers, the American Institute of Mining 
Engineers and the United Engineering Society 
have a joint library consisting of their 
individual collections. This library is located 
on the two upper floors of the Engineering 
Societies Building. The library is conducted 
as a free public library of reference, and now 
contains about 00,000 volumes, and over 
000 sets of periodicals. 



One of the most important features of the 
library is the research department, which 
places the facilities of the library at the 
disposal of out-of-town members. Upon 
application to the library, bibliographies are 
prepared on any desired engineering subject, 
and abstracts, translations or photographs are 
furnished. A small fee is charged for tnis 


A Modern Army in the Field, by Major Shipton, 
U. S. A. 

The meeting of February 3rd was attended 
by about 390 members and guests. The 
speaker of the evening was Major J. A. 
Shipton, U.S.A., Commandant at Ft. Terry, 
New York. His subject was the Conduct of 
a Modern Army in the Field. 

The structure of an army division was 
described, this being the smallest unit com- 
plete in itself, containing elements of all 
branches of the service; viz., infantry, 
artillery, signal, engineering, aerial, hospital. 
This unit contains 22,000 men, and 750 
officers. Details were given in order to 
bring out the definiteness of the structure 
and purpose of the various elements. The 
manner of maintaining communications with 
the base by means of commercial railways, 
military railways and wagon trains was 
illustrated. The completeness and defi- 
niteness of the organization were particularly 
impressive, as illustrated by the specific 
duties of the various officers and groups in 
the general structure of the army. 

Next methods of issuing orders were 
described and here again the absolutely clear 
cut manner of issuing commands by five 
paragraph typewritten orders added further 
to the impression of the absolutely methodical 
manner of conducting military operations. 
Finally the manner of conducting the army 
at the time of an offensive engagement was 

The lecture maintained the interest of the 
whole attendance and was very instructive. 
Numerous lantern diagrams were used to 
make the points under discussion clear. 

Theories of Electricity and Matter, by Professor 

The lecture by Professor Comstock, of the 
Massachusetts Institute of Technology, on 
January 6th, so interested the membership 
that in response to numerous requests a 
special course was arranged for. On Tuesday, 
February 9th, the first of the series of four or 

five weekly lectures on Modern Theories of 
Electricity and Matter was given by Prof. 

Lectures for the Near Future 

On February 17th, the Lynn Section 
listened to a most interesting talk on the 
Characteristics and Uses of Storage Batteries, 
by Mr. J. Lester Woodbridge, Chief Engineer 
of the Electric Storage Battery Company. 
The talk was illustrated by apparatus 
especially arranged for demonstration pur- 
poses. A more detailed statement will 
occur in the next issue. 

On March 3rd, Mr. A. G. Davis, the head 
of the Patent Department, General Electric 
Company, will speak of the Relation of 
Patents to Industrial Progress. 

On March 17th, Dr. W. P. Davey, of the 
Schenectady Research Laboratory, will speak 
on Recent Development in X-Ray Work. 

Electric Waves, by Professor Franklin 

Announcement was made in the February 
Review of the Paper by Prof. W. S. Franklin, 
of Lehigh University, on the subject of 
Electric Waves, given January 7th. 

The lecture covered the ground of Professor 
Franklin's recent Institute paper on Line 
Surges, but especial attention was given to 
the underlying mathematics of that paper. 
Indeed the primary object of the lecture was 
to illustrate the use of differential equations 
in physics by setting up the differential 
equations of wave motion and integrating 
and interpreting them in their application 
to some of the simplest transmission line 

The lecturer pointed out the two cases in 
which the differential equations of wave 
motion on a transmission line are integrable 
in finite terms, namely: (a) the case in which 
wire resistancejmd line leakage are zero; and 
(b), the case in which voltage and current are 
assumed to be everywhere harmonic and 
synchronous. The latter case leads to the 
ordinary problem of the alternating current 
transmission line in its steady state, and 
the former leads to an approximate solution 
(approximate because of the neglect of wire 
resistance and leakage) of the problem of 
transient effects on a transmission line. The 
lecture was devoted entirely to the latter 

Theories of Electricity, by Dr. Langmuir 

Dr. Irving Langmuir, of the Research 
Laboratory, Schenectady, read a paper on 



Friday. January 29th, on Modem Theories of 

Dr. Langmuir first sketched the historical 
conceptions of electricity and matter, and 
gradually led up to the atomic theory of 
electricity and the electron theory of the 
constitution of the atom, by means of which 
theories he explained the modern ideas of the 
conduction of electricity through gases and 
metals. The operations of the Cathode Ray 
Tube and the ordinary and Coolidge types of 
X-ray tubes were explained. 

Curves were shown of the radiation of 
energy from black bodies, and an explanation 
given of Planck's Law and the Quantum 

The various theories of the structure of the 
atom were touched on, as well as the results 
of the study of the spectra of the elements by 
means of high frequency. 

Finally, Dr. Langmuir showed how, by 
means of the new theories, many phenomena 
formerly obscure were now explained, how,the 
periodic tables of the elements have been 
supplemented and given increased importance; 
also how the gaps have been filled so that a 
continuous relation has been found between 
waves of all frequencies, from the long 60- 
cycle waves to the extremely short X-rays. 

The lecture was illustrated by numerous 
lantern slides and experiments. 

X-Rays, by Dr. Coolidge 

Dr. W. D. Coolidge, Assistant Director of 
the Research Laboratory of the General 
Electric Company, addressed the meeting of 
the A.I.E.E. January 19, 1915. in the audi- 
torium of the Edison Club, Schenectady, 
X. Y.. on the subject of Recent Developments 
with X-rays. The following is an abstract 
of his valuable address: 

Prof. W. C. Rontgen of Wurzburg, Bavaria, 
suspected that when a current of electricity passed 
through a glass tube containing a gas at very low- 
ire, invisible light waves were given off.' The 
idea occurred to him that such rays might affect a 
fluorescent screen in much the same manner as did 
ultra-violet rays. In order to cut out the visible 
light from his vacuum tube he wrapped it in heavy 
black paper. Upon operating the tube to make 
certain that the covering was completely light- 
tight, he noticed to his surprise that the fluorescent 
screen which he had left on the table three or four 
■ - away glowed brightly. 
r Rontgen investigated the properties of the 
X-rays _with characteristic German thoroughness. 
By 1897 he had amassed such a volume of infor- 
mation about X-rays that nearly "every essential 
piece of research on their properties up to 1908 can 

be found in its more elementary form in his three 
original memoirs. 

Rontgen's original tube of 1895 was, judged by 
modern standards, a pretty crude affair. The 
cathode was flat and emitted a diffuse bundle of 
cathode rays which, upon hitting the glass at the 
far end of the tube, produced X-rays. In 1896 
Campbell-Swinton added a platinum target upon 
which the cathode stream hit. This increased the 
penetrating ability of the rays obtained. In the 
same year Jackson made the cathode concave so as 
to focus the cathode stream upon a small area of the 
target. By giving more nearly a point source of 
X-rays this increased the clearness of radiographs 
tor diagnostic purposes. The X-ray tube was soon 
changed in form but not in principle. A device was 
added by which the pressure inside the tube could 
be increased at will, and various means were tried 
for removing heat from the focal spot of the target. 

Meanwhile in 1912, Dr. Coolidge discovered the 
process of making ductile tungsten such as is used in 
the filaments of mazda lamps. Shortly after this 
discovery he became interested in perfecting a 
wrought tungsten target for X-ray tubes. During 
this work it became necessary to operate the tubes 
up to the limit of their capacity in order to find out 
how much abuse the tungsten targets would stand. 
During the course of this work he found that the 
ordinary aluminum cathode could be melted if 
sufficiently high currents were sent through the 
tube. He tried to remedy this by substituting a 
cathode made of tungsten whose melting point is 
very high. But such tubes were found to be very 
unstable. When current was sent through such a 
tube, the vacuum increased rapidly until finally no 
current would pass through the tube until gas 
had been liberated from the vacuum regulator. 
From a practical standpoint such a tube was hope- 
lessly unsatisfactory. Finally it was found that if 
the process of operating the tube and immediately 
reducing the vacuum were repeated rapidly enough, 
the cathode became hot enough to glow, and that 
after this the tube would operate for several minutes 
at a time without it being necessary to let in fresh 
gas from the regulator. This suggested the idea of 
a cathode heated by some external means. 

Richardson, and others in 1902, had shown that 
electrons could be obtained by merely heating the 
cathode, but had not been able to obtain constant 
results. Dr. Langmuir, of the Research Laboratory 
of the General Electric Co., had shown that the 
rate of emission of electrons from a hot tungsten 
cathode in a very high vacuum depended only 
upon the temperature. 

If we heat a tungsten filament, electrons are given 
off and soon a condition of saturation occurs around 
the filament. If the filament is made the cathode of 
a low-potential circuit, a small current passes. If 
the voltage is increased, a larger current passes. 
Finally a voltage is reached which sweeps away 
every electron as fast as it emerges from the hot 
tungsten. For all voltages above this, the current 
is constant, and is independent of the voltage. 
Thus we have a resistance as far removed from the 
ordinary Ohm's law resistance as possible. This 
is not because the conduction is carried on in any 
different way, but because the number of available 
electrons is limited. (The reason that Ohm's law 
holds in conduction through wires is that the supply 
of available electrons in the wire is practically 

As a source of electrons in his tube, Dr. Coolidge 
made use of a small spiral of tungsten wire heated 


white hot from a storage battery in exactly the effects of electrolytes and charged colloids 

same way in which electric automobile lights are . „ • j •■;, ,, ., ? . , . 

operated. This spiral is the cathode and a block of were reviewed, with the idea of pointing a 

gas-free tungsten is the anode. The rate at which possible way for further study of the reactions 

electrons are given off from the spiral depends of immunity which seem to occur usually, 

upon its temperature, which is under the immediate if not always, in the blood between colloidal 

control of the person operating the tube. The „ <.„ n c j.r.1 ~ 1 <-• <t> -j r 

voltage across the tube is also controllable at will. P artS ° f tte solution. To give some idea of 
As the voltage employed in ordinary X-ray work tne . complexity of this field, the entire group 
is much greater than is necessary to snatch all the of immunity reactions were reviewed. These 
electrons across from cathode to anode as fast as included production of anti-toxins, precipitins, 
they are evaporated irom the filament, even at the „„„i,.4.:„:„„ u i • i i_ 1 • j 
highest currents now in use in X-ray work, the agglutinins bactenolysins, hemolysins, and 
voltage and current passing through the Coolidge a review of phagocytosis. While the cases of 
tube are totally independent. Both may be adjusted visibly complicated and augmented pha- 
to any desired value with any degree of precision gocytosis are Certainly not yet explained by the 
desired and at any such adjustment the X-ray s i mt) 1 e electrostatic "reactions nf nnrplv in 
performance of the tube can be duplicated time slm P le electrostatic reactions ol purely ill- 
after time. organic colloids, the assumption that the 

specific nature and neutralizing process of 

Lantern slides were shown illustrating the colloidally suspended toxins and anti-bodies 

development of the X-ray tube, and the might be due to different magnitudes of 

kind of work which it is possible to do with electric charges is worth considering. These 

an X-ray tube. Many of these pictures are at present explained by the assumption 

have already been published in the General of countless specific and different chemical 

Electric Review, August, 1914, and January, compounds which are assumed to be normally 

1915. present in all blood, to slight extent in all 

cases of immune or anti-bodies, and to be only 

Chemistry of the Blood, by Dr. Whitney. augmented by the process of immunization. 

On the evening of February 2nd, at the 

auditorium of the Edison Club, Dr. W. R. G - E ' Mevin e Pictures 

Whitney, Director of the Research Laboratory, On February 16th, Mr. F. C. Bateholts 

General Electric Company, delivered an delivered a lecture on the General Electric 

interesting lecture on The Physical Chemistry Company's Moving Pictures for the Panama 

of the Blood. The attendance was quite Exhibition. Mr. Bateholts displayed the 

large, including a number of medical men educational-advertising motion pictures of the 

who attended as guests. A lively discussion General Electric Company, and gave an 

followed the lecture. Among those who interesting talk on the educational and 

engaged in the discussion were W. L. R. advertising value of motion pictures. He also 

Emmet, J. B. Taylor, Dr. W. L. Towne gave an interesting account of the General 

and Dr. Krida. Electric Company's lecture bureau service. 

Dr. Whitney's talk was a review of the Among the pictures shown were those of 

properties of the blood, (with the intention of the Schenectady Works, Lynn Works and the 

showing some possible advantages to be Harrison Works, and also some pictures of 

gained by application of facts of physical the work on the Panama Canal, 
and inorganic chemistry to such a complex 

solution. It was pointed out that the blood Program for March 

is all kinds of a solution : gaseous, electrolytic, During the month of March the following 

simple osmotic, colloidal, and crass suspensoid. speakers will deliver papers in the auditorium 

Through them all the electric effects of salt D f the Edison Club, viz: 

ions and charged colloidal particles were March 2nd, W. L. R. Emmet, on Driving 

called to mind by illustrations. The charac- Ships' Propellers. 

teristic effects of the sodium and calcium March 16th, S. B. Paine, on a subject to 

ions in true blood and in physiological be announced later. 

salt solutions in case of excised hearts, and March 30th, C. D. Knight, on The Prin- 

the activity of white corpuscles in pha- ciples and Systems of Control for Electric 

gocytosis, were referred to. The mutual Motors. 





The term "noble" has been conferred upon 
certain rare gases of comparatively recent dis- 
covery, for the same reason that this term was 
long ago applied to certain metals such as gold 
and platinum, indicative of their scarcity and 
value, together with their general permanency 
under ordinary conditions. The noble gases 
now known are argon, helium, neon, krypton 
and xenon, all of which appear to be absolutely 
permanent, or chemically inert in respect to 
one another and to every other element. 


This is the most plentiful of all the noble 
gases, being present by volume in atmos- 
pheric air to the extent of about 0.9 per 
cent. Nitrogen can be made to combine 
with oxygen by the electric spark, but argon, 
being without chemical affinity, cannot be 
thus oxidized. Because of this peculiar lack 
of affinity, argon was originally obtained by 
Cavendish in 17S5 as a permanent gaseous 
residue after the complete oxidation of atmos- 
pheric nitrogen by the electric spark and the 
absorptionincaustic potash of the products thus 
formed. Cavendish, however, did not recognize 
his residue as a new element and his interesting 
experiment bore no fruitful results until it was 
repeated in 1S94 by Rayleigh and Ramsay. 
These distinguished scientists discovered its 
elementary character by an examination of its 
spectrum, and it was named "argon" from 
two Greek words signifying "without work," 
i.e., without chemical affinity, for it has been 
found impossible to produce a combination of 
this gas with any other element. 

Unlike the other noble gases, the lumines- 
cence of argon under electrical excitation in a 
vacuum tube is feeble. It is chiefly remark- 
able for a change in color from red to blue 
according to the density of the exciting cur- 
rent; a weak current producing a red lumi- 
nescence, which changes to a blue when the 
current density is increased. 

The spectrum of argon consists of many 
lines extending throughout the visible range, 
and the change in color of luminescence from 
red to blue is chiefly caused by a strengthen- 
ing or weakening of the red and blue lines 
respectively, so that either one or the other is 
presented to the unassisted eye as the pre- 
dominating color of the light. 

on is a monatomic gas, which signifies 
that its atom and molecule are identical. It is 
19.94 times heavier than hydrogen, its atomic 
weight being 39. SS, considering oxvgen as 16. 


A study of the spectrum of the sun's corona 
by Lockyer in 1868 revealed a bright yellow 
line which could not be found in the spectrum 
of any other element known at that time. The 
unknown clement which produces this line 
was named "helium" (from the Greek work 
for sun) and it remained a puzzle to scientists 
until 1895, when Sir William Ramsay observed 
the same bright yellow line in the spectrum 
of a gaseous mixture extracted from the rare 
mineral cleveite. He finally succeeded in 
isolating an elemental gas from this mixture, 
the spectrum of which showed many very 
beautiful lines, prominent among which was 
the brilliant yellow line that had been first 
detected by Lockyer. This line is known in 
spectrology as D3 and its wave length is 
approximately 5S75.5 Angstrom units (1 
A.U. = 10- s cm.). 

The complete spectrum of helium includes 
seven principal lines, colored respectively 
red, yellow, green, blue-green, blue, blue- 
violet and red-violet, but the vivid brilliancy 
of the yellow line D3 gives a strong pre- 
dominating yellow tone to the luminescence 
of this gas when it is confined in a capillary 
tube and excited by electricity. 

Helium is the second lightest of all the 
gases, its specific gravity being 3.99 as com- 
pared with oxygen at 16. It is particularly 
remarkable in being a bi-product of the dis- 
integration of radium, and therefore a strik- 
ing example of the natural evolution of matter 
from one apparently elementary state to 
another, or, in other words, of the transmu- 
tation of the elements. This gradual forma- 
tion of helium from radium is, however, the 
result of interatomic energy over which we 
have at present no control, either to hasten 
or retard its operation; so it cannot properly 
be cited as a modern realization of the ideas 
of the ancient alchemists who imagined the 
possibility of transmuting one element into 
another by a chemical process. 

Helium is monatomic, and like its com- 
panion noble gases, it shows no affinity for 
any other element, no compounds of helium 
having been discovered. It has been extracted 
from other minerals besides cleveite and has 
also been found in certain mineral waters, 
notably in the hot water from the King's 
Well in Bath, England, in which it is asso- 
ciated with argon. It is also present in the 
atmosphere in very minute quantity. 

W. S. Andrews. 



The purpose of this department of the Review is two-fold. 

First, it enables all subscribers to avail themselves of the consulting service of a highly specialized 
corps of engineering experts, or of such other authority as the problem may require. This service provides 
for answers by mail with as little delay as possible of such questions as come within the scope of the Review, 

Second, it publishes for the benefit of all Review readers questions and answers of general interest 
and of educational value. When the original question deals with only one phase of an interesting subject, 
the editor may feel warranted in discussing allied questions so as to provide a more complete treatment 
of the whole subject. 

To avoid the possibility of an incorrect or incomplete answer, the querist should be particularly careful to 
include sufficient data to permit of an intelligent understanding, of the situation. Address letters of inquiry to 
the Editor, Question and Answer Section, General Electric Review, Schenectady, X. }'. 


(130) When two similar transformers are connected 
for changing two-phase current to four-wire three- 
phase, what are the percentages of the windings 
included between the taps, and what are the 
vector relations of the current and the voltage at 
unity power-factor and at less than unity power- 

Referring to Fig. 1, which shows the connections 
for -transforming four-wire two-phase to four-wire 
three-phase by two transformers, b is a 50 per cent 
tap on the secondary winding of one transformer, 
and c and / are and 86.7 per cent taps respec- 
tively on the secondary of the other transformer. 

Two -phase 


vwwwwV'9 Cwwwww^ 

b e f 

#/^A/wywv\/\c• ^A/wwvwwg 

-£ H 

? Neutral 

Fig. 1 

Fig. 2 represents the current and the voltage 
vector relations in the primary windings, and Fig. 
3 the same in the secondary windings. In the vector 
diagrams, the voltage and current vectors at unity 
power-factor are represented by the solid lines, 
while the dotted lines indicate the position of the 
current vectors when lagging at an angle of 6. The 
direction of the vector is based upon the assumption 
of the positive direction of the currents bem? as 
shown by the arrows in the transformer windings 
in Fig. 1 R-K.W. 










Fig. 2 
Two-phase vectors 

AB=CD Voltages 

I =11 Currents 

Fig. 3 
a h=bc=0.5E 1 
Jf = 0.867E , Vnltaees 
de=0.289E voltages 

ac=cf = oj = E J 
1 =2=3 Currents 




(131) Please explain the reason for the following 

A bank of three, 6600/220-110 volt, single- 
phase transformers were connected with both 
their primaries and their secondaries in delta. 
One of these units was of 4 kv-a. and the other 
two of 3 kv-a. Wishing to obtain single-phase 
power to operate a contactor board, the discon- 
necting switches 1 and 3 in the primary leads of 
the transformers were closed (see Fig. 1). Not 
having an a-c. voltmeter, a 250-volt lamp was 
placed across the secondary of the 4-kv-a. trans- 
former, but it did not burn at 220-volt brilliancy. 




3 A <* \\ 3 /OV. \ -4 KtV. 


AA/vWVW\ /vWvW, . A/vWvWWW 
Trarts. |_l Trans. |_l Trans. 

6600 fo/ts 

220 yolts 



Fig. 1 

After about five minutes operation as just 
described, the two 3-kv-a. transformers began to 
smoke profusely and therefore all disconnecting 
switches were opened. The two 3-kv-a. trans- 
formers were then removed and the 4-kv-a. 
unit was thrown across the line. The lamp 
when placed across the secondary of this trans- 
former burned brightly, and no further trouble 
was experienced. 

With such a connection as is shown, whatever 
happened when the single-phase was connected 
would also have happened if the three phases had 
been connected. The action taking place, therefore, 
indicates that there must have been some mistake 
in the connections of the delta which produced a 
circulating current of sufficient magnitude to heat 
mail transformers far above their rated tem- 
perature rise. It is possible that the polarities of all 
three transformers were not the same, and that, in 
connecting es in delta, a large unbalanced 

voltage was obtained. With the connections as 
given, the two 3 kv-a. transformers are in series 
with each other, and both in multiple with the 4 
kv-a. transformer. If the polarities of the 3 kv-a. 
units are alike and the same as that of the 4 
kv-a. nothing extraordinary could happen, but if 
the polarity of the 4 kv-a. were opposite to that of 
the two 3 kv-a. transformers the machines of the 
whole bank would add their voltages in series and 
the circulating current in the delta would be limited 
only by the sum of their three impedances. A heavy 
circulating current would also result if the two 3 
kv-a. machines were unlike polarity, but con- 
nected as though they were alike. 

A possible difference in the ratios of the three 
transformers would also give an unbalanced voltage 
on the secondary side, which would cause circulating 

With connections as shown in Fig. 1, the trouble 
was no doubt due either to a difference in polarities 
or ratios. R.K.W. 


(132) Why does running single-phase produce such 
disastrous results in a three-phase alternator? 

Doubtless the effects referred to in the question 
were noted in a solid steel rotor alternator and 
were made known by the iron of the field becoming 
seriously overheated. 

The reason why such an action may take place in 
a three-phase alternator under the conditions named 
will be made obvious by a comparison of the behavior 
of the flux in a polyphase alternator with that in a 
single-phase alternator. 

In a polyphase alternating-current generator the 
armature reaction (or the magnetomotive force of 
the current) is constant in intensity, and revolves 
synchronously with regard to the armature, i.e., 
stationary in relation to the field. 

In a single-phase alternator the armature reaction 
is pulsating and ranges between zero and n I\/2 
(«=the number of turns per pole and 7 = the cur- 
rent per turn in effective amperes). The flux through 
the field poles, which is the resultant of the constant 
field excitation and the pulsating excitation of the 
armature, is a pulsating flux of twice the frequency 
of the machine. Consequently, in the field of a 
single-phase machine there would be the heavy 
hysteresis losses corresponding to double the fre- 
quency of the machine, if some means of lessening 
them was not employed. To accomplish this pur- 
pose, a laminated construction is used in the field 
of single-phase machines or heavy damping windings 
are provided and arranged in such a manner as to 
cut down the pulsating effect of the armature 
current. In those single-phase machines where the 
hysteresis losses would tend to be particularly 
excessive, due to the flux pulsation, both of the 
remedial measures named are embodied in the 
design of the machine. 

It will readily be seen from these descriptions 
that the standard polyphase' alternator cannot be 
expected to operate satisfactorily as a single-phase 
machine except at a considerably reduced output. 


General Electric Review 


Manager, M. P. RICE Editor. JOHN R. HEWETT Associate Editor. B. M. EOPF 

Assistant Editor. E. C. SANDERS 

Subscription Rates: United States and Mexico, $2.00 per year: Canada, $2.25 per year; Foreign, $2. SO per year; payable in 
advance. Remit by post-office or express money orders, bank checks or drafts, made payable to the General Electric Review, 
Schenectady, N. Y. 

Entered as 3econd-class matter, March 26, 1912, at the post-office at Schenectady, N. Y., under the Act of March. 1879. 

VOL. XVIII., NO. 4 tyGeZT&UcLpany APRIL, 1915 


Frontispiece .... . . 230 

Editorial: The Paths of Progress . . 231 

The Status of the Engineer . ... 234 

By Dr. E. W. Rice, Jr. 

The Absolute Zero, Part II ... 238 

By Dr. Saul Dushman 

Operating Conditions of Railway Motor Gears and Pinions . . ... 249 

By A. A. Ross 

X-Rays, Part I .... .258 

By Dr. Wheeler P. Davey 

The Modern Mine Haulage Motor ... 264 

By C. W. Larson 

The Eye and Illumination 268 

By H. E. Mahan 

The Fort Wayne Electric Rock Drill 273 

By C. Jackson 

Some Notes on Induction Meter Design . . . 277 

By W. H. Pratt 

Sign and Building Exterior Illumination by Projection . 282 

By K. W. Mackall and L. C. Porter 
Electrophysics: Application of the Electron Theory to Various Phenomena 287 

By J. P. Minton 

Railway Motor Characteristic Curves .... • 296 

By E. E. Kimball 

The Osborn Electriquette .299 

By O. E. Thomas 

Notes on the Activities of the A. I.E. E. . .301 

Practical Experience in the Operation of Electrical Machinery, Part VII ... 304 

Imperfect Slip-ring Contacts; Equalizer on the Wrong Side; Generators Motoring 
at No-load; Changing Motor Mounting; Motor Throwing Oil 

By E. C. Parham 

From the Consulting Engineering Department of the General Electric Company . 308 


Question and Answer Section . . • ■ 

The Coohdge X Ray Tube and Accessories. At the left is the tube inside a lead-glass" bowl, both 

mounted on a stand so as to be easily adjusted to any convenient position and angle. 

At the right on an insulating stand are the storage battery for heating 

the cathode filament and the rheostat by which the 

temperature of the filament is adjusted 


When the full text of the papers read on 
"The Status of the Engineer" at the mid-year 
convention of the A.I.E.E. in New York on 
February 17th, is published in the Proceed- 
ings of the Institute it will undoubtedly 
attract wide attention and lead to consider- 
able discussion, as the subject is one of 
paramount interest to every unit of the 
profession. We reproduce the very interest- 
ing paper read by Dr. E. W. Rice, Jr.. 
President of the General Electric Company, 
and feel sure that our readers will derive 
considerable pleasure from the manner in 
which he has handled the subject. 

At present we do not propose to enter into 
a discussion of the specific remarks in these 
papers, but rather to make a brief analysis of 
the general subject. 

There has been, and we feel with perfect 
justice, a very general feeling that the engineer 
has not received due recognition or com- 
pensation for the part he has played in 
remodelling our state of civilization. We 
also feel that every one who has anything 
like an adequate conception of the changes 
that have been made both directly and 
indirectly by the activities of the engineer 
will fully reciprocate in this feeling. If we 
concede without argument that this is the 
case, the interesting problem would seem to 
be to ascertain the whys and wherefores of 
this condition of affairs. We confess that 
the problem is not particularly easy, but there 
seem to be certain factors which upon analysis 
are fairly evident. To make our point, a 
brief mental picture of "what was" and 
"what is "in one particular instance will help. 

Let us consider the old warship of a hun- 
dred years ago and compare it with 
the latest modern superdreadnaught, and 
we must remember these old men-of-war 
aroused admiration and respect fully akin 
to that accorded to the latest leviathans. 
One hundred years ago the finest ship afloat 
absolutely depended upon the wind for her 
motive power and if becalmed was of little 

more use than a log floating on the surface 
of the waters. Her armament consisted 
of cast iron muzzle loading guns which were 
laboriously handled by manual labor and 
there was nothing in her from stem to stern 
that resembled a machine with the possible 
exception of the pumps, composed of hollow 
tree trunks, and the capstan that raised the 
anchor in the good old fashioned way. 

Such is the ship "that was" and what a 
contrast to the ship "that is"! The modern 
superdreadnaught with engines of 60,000 
horse-power, and rumor has it that at a 
pinch with forced draft 100,000 horse-power 
is not impossible. A speed independent of 
wind and weather of 25 knots and may be, 
if necessary, this can be increased to nearer 
30 knots per hour. Batteries of 15-inch guns 
that could hurl projectiles weighing nearly 
a ton for distances not far short of 20 miles. 
Torpedo tubes capable of firing torpedoes 
21 inches in diameter and every vital part 
protected with solid steel 14 inches in thick- 
ness. The modern battleship is indeed a 
most highly developed organism with man- 
made organs. 

Now all these developments that have 
changed the ship of yesterday into the ship 
of today have been brought about by the 
engineer, but it should be noted that in 
talking of the engineer we use this term 
throughout in the same broad sense as it is 
used by Dr. Rice in his address. " * * * we 
do not propose to limit our definition of 
'engineer' to one educated in or following 
the strictly technical professions of civil, 
mechanical and electrical engineering, but 
shall include in addition all educated men 
laboring in the broad fields of chemistry, 
physics, medicine and other organized sci- 
entific activities." 

With a picture of the ship of a hundred 
years ago and of the ship of to-day in our 
minds we are in a position to state certain 
facts that are vital to our analysis : 

There was no engineer on the old ship and 
when the Admiral gave his word of command 



he was telling others to perform operations 
any one of which he would have been perfectly 
capable of performing himself, and there 
was not one technicality with which he was 
not well versed. The ship was navigated, 
fought and handled in every respect by 
manual labor. 

Today the Admiral can neither "go ahead" 
nor "go astern," nor turn his ship, he cannot 
bring his ammunition to his guns, he cannot 
train or fire his modern monsters, he can- 
not handle his torpedoes, he cannot even weigh 
his anchor without his engineer. This is the 
triumph of the engineer, but it must be noted 
that the Admiral still gives the word of 
command and it should also be noted that 
it is only in comparatively recent years that 
the engineer has been ranked as an officer. 

This same general condition exists in all 
of our industries, in railways, in lighting and 
power stations, in mining, in our manufactur- 
ing industries, in fact, in everything that is 
modern and up-to-date. We owe the incep- 
tion, development, and successful operation 
to the engineer. The engineer is the one 
indispensable factor, and without him and 
his work progress would be at a standstill; 
and yet he neither controls them nor dictates 
the policies of the thing of his own creation, 
and what is of more importance he apparently 
is constantly seeing others reap the harvest 
for which he has so diligently sown. Now 
we believe that such is the case, but we know 
that no effect is produced without a definite 
cause, so the cause for this state of affairs is 
the interesting thing: 

Firstly, we believe that there is one fact 
that is often lost sight of, namely, that when 
the engineer is really successful, in the 
material sense, besides doing what, for 
want of a better term (and to avoid the 
misnomer of pure engineering), we shall call 
engineering work proper, he becomes active 
in organizing the work of others, and 
if successful in his broader activities his 
work often leads him further and further 
from engineering problems and more and 
more to organizing and managing the work 
of others until he is recognized as manager 
or head in his particular sphere of work. 
As he assumes wider responsibilities his 
financial reward increases, but he is not then 
the highly paid engineer, but the successful 
manager or business man, as the case may be. 
In these cases, and we believe they are many, 
the engineer has reaped the harvest for which 
he has sown, but not necessarily as an engineer. 
So when the engineer gets to the stage of 

giving the "word of command" he is often 
lost sight of as the engineer and assumes 
another title. 

Again, there are many fields of engineering 
activities where the engineer has done his 
work so perfectly that the very work that 
required engineering talent in the past has 
been reduced to almost routine work and 
can now be successfully performed by the 
machines of his creation and a type of labor 
partially or totally unskilled. In such cases, 
and they are legion, the engineer has displaced 
himself by the product of his own brains. 
Indeed, in the field of operating engineers 
this is particularly noticeable, where the 
perfection of mechanical and electrical devices 
has been brought to such a state that the 
engineer holds much the same position as a 
lifebelt — for the greater part of the time he 
is not wanted, but when he is wanted his 
services are imperative if a disaster is to be 
avoided. This situation has reduced the 
number of engineers employed in the operation 
of large engineering undertakings, and we 
often find only one or two engineers directing 
the work of a host of less skilled attendants 
from whom no great degree of technical 
knowledge is required. 

So it would seem that the engineer has in 
a multitude of cases displaced other workers 
by introducing new methods and again by 
the perfection of his own devices in turn 
displaced himself. This undoubtedly has led 
to many men of good engineering training 
having to perform work which those with a 
less costly preparation could perform almost 
as well, which naturally leads to dis- 
satisfaction. This is most unfortunate for the 
would-be engineer, but it seems inevitable 
and apparently is the same in some other 

So in the field of operation and construction 
we shall still see the old rule of life prevail — 
many will enter this field of activities but 
comparatively few will become really suc- 
cessful in the material sense — and we shall 
still see many men discontented with their 
lot not necessarily because their work is not 
congenial, but rather because after an expen- 
sive education and much self-sacrifice and 
arduous labor in early life they are not able 
to reap the harvest they feel that they have 
sown for. 

The real field for engineering is the same 
today, and will be in the future, as it has 
been in the past, namely, development work, 
showing the world at large "how to do for 
one dollar what a fool can't do for two," 



and it is to this great field of development 
work that we must call the most able young 
men of this generation and of generations to 
come ; our future absolutely depends upon the 
engineer in just the same degree as our past 
progress and prosperity has been due to 
his work. Anything that discourages the 
brains of future generations from wishing 
to enter the engineering profession is a menace 
to our future welfare, and it is for this reason 
that we feel apprehensive of any movement 
that would make the engineering field appear 
less attractive. 

If the feeling generally prevails that 
engineering work is becoming so standardized 
and so reduced to routine work that it is not 
worth a young man's while to prepare himself 
at the great cost involved for the profession, 
or again if he should feel that , even if he were 
fortunate enough to work himself up to a 
position where he was really doing important 
development work, that the reward would be 
altogether inadequate for the effort he has 
expended then we are not going to progress in 
the future as in the past. The spreading of this 
feeling must be avoided. We fully recognize 
that our future economic stability demands 
the organization of engineering work and that 
it is essential after developments have been 
made by the engineer that production and 
operation must be standardized as far as 
possible, and indeed, that this is one impor- 
tant phase of the engineer's work; but we also 
realize that if the idea prevails that this 
organization is being pushed beyond its 
limits to the extent that it is inimical to the 
status of the engineer we shall have many 
difficulties to face in the future. It seems 
that we should certainly form our policies 
in such a way that the engineering profession 
shall never come to be regarded in the same 
light as journalism, of which it has been so 
often said that it is an excellent profession 
to get into if you are quite sure you can 
get out. 

There is another side of the question; and 
one of the speakers in New York thought 
the engineer had nothing to complain of 
and that all things being considered the 
average engineer was as well rewarded for 
his work as the average man in other pro- 
fessions. This would be hard to prove or 
disprove without a most exhaustive study, 
and even if it were proved, the point would 
still remain that the engineering profession 
is giving more to the world than any other 
profession, and it is essential that it should 
be attractive to the voung man of the future. 

Certainly there are many walks of life in 
which the material rewards are all out of 
proportion to the service rendered to the 
state when compared with those in the 
engineering profession. A young man in 
choosing his profession naturally realizes 
this, but in the recent past the engineering 
professions have been talked of as those of 
the greatest possibilities, so that if the feeling 
becomes general that these possibilities are 
not as good now as they were in the past we 
shall fail to secure the most desirable young 
men of today as our engineers of the future. 

Up to this point we have only talked of 
material rewards and now if we regard the 
engineering profession in another light it 
seems that the rewards are far above those 
in most other professions. All those engaged 
in the great modern science of development, 
to which our engineering professions have 
been so largely reduced, have the immeasur- 
able joy of achievement or the possibility of 
achievement, and it is the intense interest in 
striving for accomplishment that makes the 
engineering professions what they are, and 
what has made the engineer the man of 
courage and resourcefulness, of patience and 
determination, of self-sacrifice and unending 
work. The very intensity of work with 
which the engineer devotes himself to his 
daily task precludes the constant thought of 
self-advancement and the desire to leave an 
all absorbing field of activities for others 
where the material reward would be greater, 
but the interest and worth-whileness of life 
would be less. 

The engineer must often have the idea 
that he is being exploited by others because 
of this very loyal devotion to work rather 
than to self-interest, and undoubtedly this 
has been the case in many instances; but we 
hope and trust that the very fact that the 
engineer has changed the world to such an 
extent that we are finding it every day more 
necessary that our commercial men, finan- 
ciers, etc., should know more of the engineer 
and of his work to enable them to transact 
business in a world whose modern foundations 
rest on a structure of engineering accom- 
plishments will lead to a more perfect under- 
standing, and, may be, to a better material 
reward for the engineer in the future. All 
of these different units have a common 
object; their work is the part of one great 
plan, and any factor in our great scheme of 
life that is so absolutely indispensable to our 
future progress must surely hold an enviable 
position in years to come. 




By Dr. E. W. Rice, Jr. 

President, General Electric Company 

The author, who has "lived with and worked alongside of engineers for more than thirty years," has written 
this contribution from his rich experience. He relates some of the achievements of the engineer and calls 
attention to the changes that this work has brought about in our state of living during the last four decades. 
The personal characteristics of the engineer are referred to in an interesting manner, and stress is laid on the fact 
that honest v is natural to the profession. Dr. Rice thinks that it is now incumbent on the engineer to take 
a hand in the greatest work of all, the government of the country, by showing an active interest in the 
framing of our laws and in guiding the work of the many commissions that form such a prominent part in our 
modern government. This address was read before the A.I.E.E. on the evening of February 17, 1915. — Editor. 

The status of the engineer is an important 
subject, and should be of vital interest to 
even - one of us. 

It is well for us to pause a few moments 
from our daily task and make a brief survey 
of the engineer's work, to consider its impor- 
tant influence upon the life of this busy 
world, and especially to enquire what new 
service awaits the engineer now and in the 
immediate future. 

During this discussion we do not propose 
to limit our definition of "engineer" to one 
educated in or following the strictly technical 
professions of civil, mechanical and electrical 
engineering, but shall include in addition all 
educated men laboring in the broad fields of 
chemistry, physics, medicine and other organ- 
ized scientific activities. 

I do not think that we can be accused of 
serious exaggeration in saying that the world 
is indebted to such men for the application 
of steam to ships, cars and workshops ; for the 
invention of the sewing machine, the type- 
writer and the phonograph; for the intro- 
duction of the bicycle, automobile and 
aeroplane; they have brought the marvels 
of photograph}' into existence, giving us the 
moving picture, X-rays and colored photo- 
graphs. High explosives have been created 
to build and to destroy. We must thank such 
men for the untold blessings of anesthetics; 
for showing us how to successfully limit and 
combat epidemics of dread diseases. 

Coming to our own special field, the 
members of our profession have given the 
world the telegraph, the ocean cable, the 
telephone and the wireless; created electric 
lights for our homes, cities and workshops; 
the electric motor to run our trolley cars, 
railroads and factories; have designed' dyna- 
mos and great transmission lines with which 
to save and make useful the otherwise 
wasted power of our waterfalls. These and 
many other contributions equally wonderful 
and equally useful — miracles at first but now 
mere commonplaces and necessities — have 
been evolved from the brains of our busy 

scientific engineers largely during the past 
40 years. 

But I will not weary you with a further 
recital of engineering achievements, as such 
a recitation of even the shortest possible 
catalogue would consume the entire evening. 

My object in thus calling attention to the 
relatively recent contributions of engineers 
to the wealth and resources of the world is 
not to tickle your pride in belonging to the 
engineering profession, but rather to awaken 
your sense of responsibility for the great 
changes in our daily life, our methods and 
opportunities of conducting business and 
all other activities, which have been 
brought about directly and indirectly by 
such accomplishments, and to make some 
suggestions for the meeting of this responsi- 

Is it not a fact that civilization in its present 
form would never have arisen and would 
speedily come to an end if deprived of the 
engineer and his services? Has not the 
equilibrium of the world been upset by these 
very gifts of the engineer? 

Is it not evident that such tremendous 
additions to our power, knowledge and wealth 
must have a powerful influence upon every 
phase of our existence? Have not our relations 
with nature and with each other been pro- 
foundly affected and as a result required 
many new adjustments? 

The discovery of new trade routes has, 
as is well known, completely changed in 
times past the history of nations and the 
fate of their peoples. The discoveries of our 
scientific engineers during the past 40 years 
have been of greater importance than dis- 
coveries in trade routes, and it is inevitable 
that in adapting itself to the new conditions 
society should be deeply affected. The 
adaptation of man to his new environment 
could not take place without strain and 
friction. Are we not now in the midst of 
such a process of adjustment? 

Of course, we all appreciate that the labor 
and thought of manv other men of vision 



and enthusiasm were needed; men experienced 
in finance, commerce, trade and government, 
to render the all essential aid required to 
introduce and to adapt to our daily lives 
these great contributions. But it would seem 
to be self-evident that without the creative 
work of the scientific and technical engineers 
these things would not have seen the light 
of day. 

This remarkable development was fairly 
started during the first half of the 19th 
century under the guidance of the civil, 
mechanical and chemical engineer, but was 
tremendously accelerated by the advent of 
the electrical engineer about 40 years ago. 
His work during the past decades has reacted 
upon that of the other engineering professions 
and stimulated and made possible the almost 
equally marvelous development in mechani- 
cal, chemical and other lines of activity. 
Therefore, I regard all those who have been 
able to participate in the service of electrical 
science as happy and fortunate individuals. It 
is true that the financial reward has not always 
been great; on the contrary, it has often been 
extremely meager when compared with the 
rewards which frequently come to the success- 
ful lawyer, financier or merchant, but our 
engineer has been rewarded by something 
more valuable and precious than gold — the 
thrilling joy of achievement. There can be 
no greater satisfaction than that which 
comes to a man who believes that he is the 
first to discover some new force or to make 
some new and useful invention. 

I have lived with, and worked alongside of, 
engineers more than 30 years. I think I 
understand the engineer's aspirations and 
character. I can say that it is a case where 
familiarity has not bred contempt, but, on 
the contrary, has inspired respect and 

The engineer is popularly supposed to lack 
certain qualities needed in a successful man 
of business, or to make a good salesman, 
or to handle important financial matters, or 
to fill positions requiring general executive 
ability. Is this popular idea justified? We 
may admit that an engineer who has devoted 
his entire time to his exacting work may be 
lacking in the knowledge and experience of 
other lines of activity, but it does not prevent 
him from having certain natural qualities, 
integrity, tact and aggressiveness combined 
with general intelligence and common sense. 
These qualities are personal and not pro- 
fessional. No group of men has a monopoly 
of such qualities and in none are they entirely 

lacking. These qualities are to be found as 
generally among engineers as among other 

It has been further charged that as an 
engineer deals with nature and natural laws 
his experience has been limited to impersonal 
objects, and that he must fail to appreciate 
or understand the complicated human ele- 
ment which is the important factor in business 
or in political life. This may be also partially 
true, particularly in the case of some of those 
whose work has been confined to that of pure 
research or pure science, but is not a general 
condition even among such men, and by no 
means the condition among engineers who 
of necessity are brought more or less in 
contact with the human element. 

I have noticed that an engineering educa- 
tion and training have generally developed 
a man's powers of observation and his desire 
and ability to learn. He becomes skeptical 
of mere theories, doubts tradition and spurns 
superstition, but he constantly searches for 
the truth and is not afraid of facts. He 
habitually tries to see things as they are and 
not as he thinks they should be. . He is never 
satisfied that "whatever is, is right," but is 
ever trying for something better. I do not 
need to tell this audience that engineers 
do not always agree as to the interpretation 
of facts, but opinion is frankly based upon 
facts and not upon preconceived notions. 
One who refuses to face or acknowledge facts 
loses his influence upon his fellows and his 
standing among his brother engineers. The 
engineer is always "from Missouri." 

There is an old proverb which runs some- 
what as follows: "One look is worth a 
thousand words." I like that proverb, and 
it is, I think, a fair description of an engineer's 
point of view. How often you hear the 
expression among engineers: "Well, let's go 
and take alook at it." Is not this the spirit 
which is needed in respect to other problems 
in the social, industrial and political world? 
Do they not need less talking about and more 
intelligent looking at? 

It is true that the engineer deals primarily 
with nature, but nature does not lie. The 
engineer, therefore, leams early in life the 
utter uselessness and folly of deceit. He 
knows that it would be silly to the point of 
insanity to try to fool nature. He is constantly 
on his guard not to fool himself and is there- 
fore not likely to try to fool others. In fact, 
he loses in time the desire to deceive, even 
if he ever had it. Honesty becomes a habit, 
not the honesty of the old line trader formu- 



lated in the saying "Let the buyer beware," 
but the kind of honesty which scorns to take 
advantage of the negligence or ignorance of 
his customer, which involves honest thinking 
as well as honest action. It is quite possible 
that this habit may make him at first the 
easy prey of dishonest men, but it is a quality 
which commands respect and which wins in 
the end. It is needed and appreciated in 
business of all kinds and sizes, little and big. 
It is helpful to little business. But big business 
is doomed to big and disastrous failure unless 
saturated with honesty. 

The engineer's training also tends to 
produce in him a fine blend of conservatism 
and radicalism. He is not afraid of a thing 
because it is new and he is not slavishly 
bound to precedent; on the contrary, he is 
frequently the creator of new things and a 
breaker of precedent, but he also believes 
in continuity and is not likely to let go of the 
old until he has a good hold of the new. 
He does not adopt an idea merely because of 
its novelty, but demands before adoption the 
acid test that it should be reallv better than 
the old. 

There is, therefore, a large field of service 
open to the engineer in manufacturing, 
commerce, farming and all other business 
activities of our country for which his 
education and training have made him 
eminently fit. In fact, his work in science 
and engineering, already briefly alluded to, 
has succeeded in so increasing the magnitude, 
variety and intricacy of manufacture and 
trade that the special knowledge of the 
trained engineer is already in demand in 
almost all departments of our commercial 
and business life. Even in the specialized 
field of selling, the old type of salesman with 
precious little technical knowledge has been 
largely displaced by the engineer salesman. 

There is, however, another opportunity for 
service awaiting the engineer of a most 
valuable and patriotic character. The biggest 
business after all is that of running this great 
country of ours. The United States not only 
operates the largest businesses itself in its 
various departmental activities, but through 
its legislators and various commissions it has 
taken a lively and paternal interest in private 
business. It makes the rules for the conduct 
of our business which fundamentally affect 
our future for good or for evil. It seems to me 
that the engineer ought to take an important 
part not only in conducting this great 
enterprise but in helping to make the rules 
for our faith and conduct. 

I recently heard a member of Congress 
say that in looking at Congress one was merely 
seeing as if reflected in a mirror the great 
people who elected it, and that if we, the 
people, did not like the looks of ourselves 
we should not get angry and break the mirror ' 
but go and wash our faces. Now, while that 
was a very humorous and witty simile it 
seemed also to convey a homely truth and a 
sensible suggestion. I began to wonder how 
much there was in the suggestion, and thought 
I would ascertain just how accurate a reflec- 
tion of our people and its activities was to be 
found in Congress and Legislature. I thought 
it would be interesting to learn the profession 
or avocation of those whom we have elected 
to represent us in Congress. I have here a 
list from which I will briefly abstract : 




Lawyers i 1 

Farming 5 

Banking 4 

Publishing 4 

Merchants, mfrs., 
railroads, real es- 
tate 7 

U. S. Navy 1 

Medical profession 1 

Not specified 3 

Total . "96 



Lawyers 275 

Editors and pub- 
lishers 23 

Merchants and mfrs. 32 

Other business 32 

Farming 14 

Banking 4 

Educational profes- 
sion 6 

Medical profession 5 

Architects 3 

Engineers 1 

Not specified 40 

Total 435 

It will be noted that 75 per cent of the 
Senators are classified as lawyers, and 65 
per cent in the House come under the same 
classification. I may say, incidentally, that 
I did not find a single one among the Senators 
who professed to be an engineer, and only 
one in the House of Representatives. An 
examination of the roster of the State of 
New York shows a similar condition, a large 
majority of the membership of both the 
Senate and Assembly being classified as 
lawyers. Now, I do not know how these 
facts impress you, but the witty simile of 
which I spoke rather lost its point as a 
conveyor of homely truth in the light of the 
facts. A body whose composition is about 
70 per cent lawyers cannot be considered as 
a very accurate reflection of the people of this 



Now, I have the utmost respect for mem- 
bers of the legal profession. We are all 
constantly trusting lawyers with our most 
important business matters and intimate 
private affairs. No profession has higher 
ideals and no profession comes nearer to 
realizing these ideals in practice. They 
deserve our confidence. I also yield to no one 
in my admiration of the ability, integrity 
and patriotism of the great men whose 
names have honored the legal profession and 
shed luster upon our country; men who 
frequently at great personal sacrifice have 
given the best part of their lives to the 
service of their country. 

However, I think it is competent for us 
to enquire as to whether there is not a dis- 
proportionate number of members of the 
legal profession in our law making bodies. 
Is it for the best interests of this country 
to have any one kind of talent and training 
or point of view so overwhelmingly rep- 
resented? There is a pretty general opinion 
in this country that we are afflicted with 
too large a number of laws, and it has been 
suggested that there may be a connection 
between the number of laws and the number 
of lawyers in our legislative bodies. Is it not 
also a strange anomaly that a country which 
owes so much of its phenomenal prosperity 
to the creative work of engineers should have 
practically excluded such men from its 
Congress and Legislatures? Would not our 
general condition have been better if years 
ago we could have injected into the composi- 
tion of our law making bodies a number of 
high class, sensible engineers?' It seems to me 
that our engineers have a duty to perform, 
that they owe it to themselves and to the 
country not to be satisfied with being simply 
hired to give their views and professional 
opinion upon programs prepared by other 
men, but should sit with our rulers and share 
directly in the responsibilities of government. 

It is reasonable to expect that men who 
have been the greatest factor in the creation 

and conservation of our material wealth 
and resources should have sound and con- 
structive ideas of practical value upon the 
matters which our commissions are created 
to control. Therefore, our great Commis- 
sions which are charged with such tremen- 
dous power and grave responsibilities 
should have among their members com- 
petent engineers of experience as well as 
lawyers, practical business men and experts 
in the special province over which the Com- 
mission has jurisdiction. 

One of the most hopeful signs of the times 
is the great awakening of the business men 
of this country to the imperative necessity 
of taking an intelligent interest in our 
Government, and it looks as if our business 
men now propose to make a business of 
seeing to it that they are properly represented 
in the business of government. Engineers 
should arouse themselves and participate 
in this great movement. 

While up to the present no better or more 
practical means has been discovered than 
our great political organizations for giving 
effect to the wishes of our citizens, it is 
becoming increasingly evident to thinking 
men that no permanent advance can be made 
by simply turning out one political party and 
substituting representatives of another as 
our rulers. An intelligent and continuous 
effort should be made to improve the com- 
position of our legislative bodies. We are 
essentially a nation of manufacturers, traders 
and farmers. We are all part of an organiza- 
tion with a mechanism which is so delicate, 
extensive and complicated that it must be 
controlled and managed with the greatest 
wisdom and intelligence if we wish to continue 
to progress in prosperity and lead happy and 
useful lives. It seems to me that in the 
future it will be the duty as well as the 
privilege of the engineer who so largely 
contributed to the production of this compli- 
cated mechanism to assist in its management 
in order to assure its preservation. 



Part II 

By Dr. Saul Dushman 
Research Laboratory, General Electric Company 

This article is a continuation of the contribution by Dr. Dushman that appeared in our February issue, 
and contains a summary of the results that have been obtained during the last few years from investigations 
on the properties of substances at extremely low temperatures. The discovery of a "superconducting" state 
for pure metals at these temperatures is especially noteworthy. The general conclusion toward which these 
investigations lead is that at very low temperatures the properties of all substances tend to obey very simple 
laws and that all these properties are probably connected by functions which are of the same form for all 
substances. — Editor. 


In a previous issue we reviewed very 
briefly the logical foundations of our present 
temperature scale and the various methods 
that have been used to attain extremely 
low temperatures. Before proceeding to 
discuss the behavior of different substances 
at these low temperatures, it may not be 
out of place to digress briefly in order to 
point out reasons which have impelled 
physicists to undertake laborious and difficult 
investigations in a field which at first sight 
might appear so "impractical." For, after all, 
we live in a pragmatic age and the layman 
may be pardoned for asking the pertinent 
question, "Of what use is it ; " 

We do not need to go further back than 
25 years to realize that for a long time 
scientific investigations were mostly con- 
fined to a very narrow region of temperatures, 
between approximately the minimum freezing 
point of ice-salt mixture, — 22 deg. Cent., and 
the boiling point of mercury, 360 deg. Cent. 
Upon the results obtained in this manifestly 
limited field were founded a number of 
generalizations and theories for the interpre- 
tation of the whole realm of natural phe- 
nomena. The fundamental principles of 
dynamics and statics, the laws of chemical 
combination, the electromagnetic theory of 
light, the classical system of thermodynamics, 
and the kinetic theory of gases — this whole 
structure was a magnificent attempt to 
explain and correlate the results of observa- 
tions in many different fields of investigation. 
It is true that the structure was somewhat 
contradictor)- in its style of architecture, 
and rather unstable in a good many places; 
but in spite of these deficiencies it appeared 
fairly satisfactory, especially if one did not 
it it as a whole, and merely considered 
the separate portions. 

But the past two or three decades have 
seen a most amazing expansion in our 
knowledge of the universe. The requirements 
of the industrial arts on the one hand, and 
the increased facilities and desire for purely 
theoretical investigations on the other, have 
both contributed to accumulate an immense 
number of observations in diverse fields of 

It was in this manner that the development 
of processes and operations involving the use 
of very high temperatures led to a more 
careful study of the properties of substances 
at these temperatures. Furthermore, it was 
necessary to devise methods of high tem- 
perature thermometry. Hence arose a number 
of investigations which finally led to a radical 
revision of all our previous concepts of 
energy. This story has been told in another 
connection, and one can only refer here 
briefly to the work of Lummer and Pringsheim 
and others on the laws of radiation of a 
black body which led Planck to formulate 
an atomistic theory of energy. The Electro- 
magnetic Theory, the Principle of Equi- 
partition, the Law of Continuity of Dynamical 
Effects — all of which are based upon the 
same fundamental equation, nay, even these 
equations, were called into question and the 
necessity arose for re-stating them in new 

There was all the more incentive for doing 
this, as discoveries in other realms of physics 
seemed to demand equally radical changes 
in our former views. The almost simultaneous 
discovery of X-rays and radioactive phe- 
nomena led to results that could not be 
correlated with the classical views. The atom 
could no longer be regarded as a metaphysical 
entity; here were atoms actually disintegrat- 
ing in front of our eyes ; we could count them 
and trace the life historv of each one. But 



during the process of disintegration these 
atoms emit sometimes positively charged 
particles of atomic dimensions, and at other 
times negatively charged corpuscles which 

possess of the mass of a hydrogen atom. 

1 sun 

Therefore, the atom itself must be a very 

complex structure. 

The theory of discontinuous emission of 
energy quanta propounded by Planck could 
thus find a parallel in the theory of radioactive 
transformations proposed by Rutherford and 
Soddy. But Planck's theory war, found to be 
capable of much further application than to 
the explanation of the laws of radiation. 
Almost immediately after Planck formulated 
his theory, Einstein pointed out that on the 
basis of the same theory it ought to be 
possible to predict the specific heats of 
bodies at different temperatures and that at 
extremely low temperatures the specific heats 
of all substances ought to decrease indefi- 
nitely. This conclusion appeared all the 
more interesting because it agreed with a 
semi-empirical conclusion at which Nernst 
had arrived from a consideration of the effect 
of temperature on the equilibrium of chemical 
and physical reactions. Since Einstein's 
deductions appeared just as valid as Planck's 
assumptions, it became of vital interest to 
determine accurately specific heats at very 
low temperatures. 

But the determination of specific heats 
has not been the only interesting problem 
in the realm of low temperatures. The 
electrical and thermal conductivity, and the 
magnetic properties are equally important 
subjects of investigations. We are still far 
from being able to apply the electron theory 
to calculate the conductivity of a metal at 
any temperature. A knowledge of the laws 
governing the variation in the electrical 
resistance of pure metals near the absolute 
zero would aid considerably in placing the 
electron theory on a more definite basis. 

The investigation of the properties of 
substances at extremely low temperatures 
thus appears of vital importance not only 
in order to refute or confirm the atomistic 
theorv of energy, but also in order to give 
us more definite views of the actual mecha- 
nism of electrical conduction in metals. 

That the results of these investigations 
are bound to profoundly affect our future 
theories of the structure of matter is quite 
evident. We have been accustomed to 
considering the gas laws as typical of matter 
in the simplest state. The kinetic theory of 

gases attempts to explain these laws by 
assuming that the molecules of the gas are 
in constant motion with an average kinetic 
energy that increases with the absolute tem- 
perature. Similarly the heat energy of solids 
is ascribed to oscillations of the atoms about 
mean positions of equilibria. From this point 
of view it would follow that near the absolute 
zero all vibrations among the atoms ought to 
decrease in amplitude considerably. Under 
these conditions might we not expect some 
general laws for solids corresponding to those 
observed in the case of gases? Attempts 
have been made in the past two or three 
years to develop a theory of the solid state 
along these lines, and while this work is as 
yet far from complete, a few generalizations 
have been deduced which are of extreme 

In the following paper we shall discuss the 
results of the investigations at low tem- 
peratures under the following headings: 

(1) Specific Heats. 

(2) Electrical Properties. 

(3) Magnetic Properties. 


The Quantum Theory 

As the quantum theory has been discussed 
very fully in another connection,* it will 
be sufficient to state rather briefly the 
fundamental assumptions of this theory and 
the reader can refer to the previous articles 
for more detailed discussion. 

In electrical engineering, we are familiar 
with the production of high frequency 
alternating currents by the discharge of a 
condenser through an inductance. We have 
in this case, a continuous oscillation in the 
electrical energy from a potential form (when 
in the condenser) to a kinetic form in the 
inductance, and the result is the emission of 
electromagnetic waves whose frequency de- 
pends upon the magnitudes of the inductance 
and capacity. Since light and heat are similar 
to those electromagnetic waves and differ 
only in possessing much higher frequencies 
we must conceive of their being likewise 
produced by some form of oscillator. In the 
case of visible light, the existence of the 
Zeeman effect and analogous observations 
lead to the belief that the radiation is pro- 
duced by the oscillation of electrons around 
positively charged centers. 

Now "the fundamental assumption made 
originally by Planck can be stated thus : In the 

*S. Dushman, General Electric Review, Sept., 1914. 



emission and absorption of electromagnetic 
energy the interchange of energy between an 
oscillator and the surrounding space can occur 
only discontinnously, in multiples of a unit 
quantum hv, where v denotes the frequency of 
the radiation and h is a universal constant. 




















' fi v -SO - 






























Fig. 1. Curves Illustrating Einstein's Formula for the Total 
Energy (£) of Solids where E =3 RF 

From this assumption Planck deduces the 
conclusion that the average energy possessed 
by anv oscillator is 

u.-X-g^- (i) 

( kT_ l 

for each degree of movability or freedom. 
In the case of a linear oscillator possessing 
two degrees of freedom the average energy 
is double, and-so-forth. 

Einstein's Formula for Atomic Heats at Low Tem- 

As mentioned above, Einstein carried 
this theory one step further by concluding 
that in the case of heat emission and absorp- 
tion there must exist similar discontinuities. 
There are good reasons for believing that the 
longer heat waves emitted by solids are due 
to vibrations of the atoms themselves. 
Moreover, consideration of the elastic proper- 
ties of solids leads also to the conclusion 
that the atoms in these cases are held together 
by quasi-elastic forces so that they vibrate 
about a mean position of equilibrium. Such 
an atom possesses therefore both kinetic and 
potential energy, is in fact an oscillator 
similar to that used for producing electro- 

magnetic waves. Einstein, therefore, felt 
justified in applying the quantum theory 
to this case and he deduced the interesting 
result that while for ordinary temperatures 
the atomic heat for most solids should be 
about six (as demanded by the Dulong and 
Petit law) this value must tend to diminish 
with decreasing temperature until it becomes 
equal to zero at the absolute zero. 

Since a vibrating atom in a solid possesses 
both kinetic and potential energy and the 
average kinetic energy must be the same as 
that of a monatomic molecule in the gaseous 
state, it follows from equation (1) that the 
average energy per atom of the solid ought 
to be 

r, = -|^ (2) 

Denoting the total energy per gram atom 
by W, and the number of atoms per gram 
atom by N (6.06X10 23 ), equation (2) becomes 

H =-tt =3 Rir. — 

e* r -l 



where k = j- r denotes the atomic gas constant, 

and /3 is written for h/'k. 

It is evident that for small values of $v 
or very large values of T, W becomes approxi- 
mately equal to 3 RT, while in all other 
cases it is less. Fig. 1 shows the form of the 

function F=- 


for different values of T 

and for the values $v = 50, 200 and 400. As T 
increases, the value of F approaches it more 
and more until at T= » the two are equal : 
but at any given temperature T, the value 
of F differs from that of T more and more as 
j3p is increased. According to the law of 
Dulong and Petit the atomic heat of all 
monatomic solids ought to be proportional to 
the temperature. This would correspond to 
the case where fiv is infinitesimal, and in Fig. 1 
it is indicated by the straight line. 

Debye's Formula for Atomic Heats 

The actual observations of Nernst and 
others on the specific heats of bodies at low 
temperatures were found to be in fair agree- 
ment with Einstein's equation at higher 
temperatures, but discrepancies became more 
and more noticeable as the temperature was 
decreased. While there was a qualitative 
agreement between the atomic heat-tem- 
perature curves calculated according to this 



equation and those actually observed, the 
formula was found to be completely inade- 
quate as temperatures nearer the absolute 
zero were approached. 

Nernst and Lindemann were therefore led 
to suggest a modification of Einstein's 
equation which gave very good agreement 
down to the very lowest temperatures. But 
their formula had no theoretical basis, and 
it was only very recently that Debye deduced 
a formula which has proven to be valid over 
the whole range of temperatures. According 
to Debye, it is not right to assume that the 
atoms of solids are capable of vibrating with 
only one frequency: The propagation through 
solids of vibrations of low frequency, such as 
sound waves, shows that there must exist a 
whole series of values of v. Only it is necessary 
to assume that there is an upper limit to this 
range of frequencies, and furthermore that 
the total number of different frequencies 
cannot exceed 3 A r . Debye shows that the 
maximum frequency (v m ) and the distribution 
of lines in this acoustical spectrum may be cal- 
culated in the case of monatomic solids from 
the elastic constants of the material, and deduces 
the following formula for the atomic heat, C v , 
at constant volume.* 

C V = ZR 

[fJV^-^1 » 

where x = 

PVm = Q. 

T T 

In obtaining this formula Debye also 
makes use of the quantum theory, but instead 
of assuming like Einstein, that each oscillating 
atom can absorb or emit different multiples 
of the unit energy quantum hv, he assumes 
that each individual frequency possesses the 
average energy 



It is rather difficult to understand exactly 
what this means physically, but we shall find 
that Keesom and others have found it 
necessary apparently to introduce somewhat 
similar ideas in order to account for the 
rate of change of the electrical conductivity 
of metals at low temperatures. 

The function 9 is designated by Debye as 
the "characteristic" temperature of the 
particular solid, and he shows that it can be 
calculated from the density, compressibility, 

* Ann. Physik. 39. 789 (1913). 

Young's modulus of elasticity and the 
modulus of torsion. 

Equation (4) may be stated thus: 

The atomic heat of monatomic solids is a 

universal function of the ratio Q/T where 9 is 

a characteristic temperature for each substance 

depending upon its density and elastic constants. 

C C 

Fig. 2 gives the plot of the value zr-= = y^- 

6 R C » 

Fig. 2. Atomic Heat of Monatomic Solids According to 
Debye's Formula 

as calculated by Debye from equation (4). 
At constant value of T/Q, the atomic heats 
of all monatomic substances are the same. 
Knowing therefore the value of 9 for any 
substance it is possible from this graph to 
determine the atomic heat at any temperature. 
For very low .values of T/Q, that is large 
values of x, equation (4) assumes the much 
simpler form : 

C„ = 3i?X77.94^ 3 (5) 

That is, at sufficiently low temperatures 
the atomic heat varies as the third power of 
the absolute temperature, or the total energy 
of monatomic solids at temperatures near 
the absolute zero is proportional to the 
fourth power of the absolute temperature. 
This relation is therefore analogous to the 
Stefan-Boltzmann law for the total energy 
radiated by a black body. 

At very high temperatures, that is ex- 
tremely small values of x, the expression 
inside the brackets in equation (2) reduces 
to unity, so that, 

which corresponds to the Dulong and Petit 



The actual observations at very low- 
temperatures are in splendid agreement with 
equation (5). In Fig. 3 are shown the curves 
for the atomic heats of diamond, lead and 
aluminum. The third-power law (equation 5) 
is actually found to hold over quite a con- 
siderable range of temperatures. Thus, in 
the case of silver it holds up to 273 deg. K., 
(0 deg. C), and in the case of diamond up 
to about 200 deg. K. Table I contains the 
values of 9 for a number of different metals. 
The data under 9i were calculated by Debye 
from the elastic constants, while those under 
02 were determined from the observations 
on the specific heat. 



fivtn =Ol 




























When it is considered that the elastic 
constants were obtained at room temperature 
and on different samples of metal from those 
used in the determination of the specific 
heats, the agreement must be considered 
as very good. 

It is interesting to note that while Debye 
deduced his formula for monatomic solids 
only, the third-power law was found by 
Eucken and Schwers* to hold just as well for 
the specific heats of the minerals fluorspar 
(17.5 deg. to 86 deg. K.) and pvrite (21.7 deg. 
(17.5 to 84 deg. K.). 

More recently Nernst has attempted to 
deduce a formula similar to Debye 's, making 
use of the assumption that the distribution 
of energy quanta takes place among groups 
of atoms, rather than among different fre- 
quencies, f But the experimental evidence is 
hardly sufficient as yet to be able to decide 
between these different theories. 

Specific Heat of Gases 

When a gas absorbs heat a part of this is 
used up as increased energy of translation 
of the gas molecules, while the other fraction 

* Verh. d. Deut. phys. Ges. 16. 578 (1913). 

t A. Eueken. Abh. d. D. Bunsen Ges.. 7. 390 (1914). 

is used up in increasing the rotational energy. 
According to the classical theory the specific 
heat at constant volume per molecular weight 
should be l /2 R=\ calorie for each degree of 
freedom. Since a molecule consisting of two 
atoms possesses three degrees of freedom 
in virtue of its translational energy and 
two in virtue of the rotational energy, the 

specific heat per molecule should be — R = 5 


Actual observation showed that at lower 
temperatures the molecular heat of hydrogen 
at constant volume tends to diminish to a 
constant value 3,+ and Nernst suggested that 
this must be due to the diminution in rota- 
tional energy as the temperature is lowered, 
just as the energy of vibration of atoms 
decreases with decreasing temperature. 

Applying equation (1) to the rotational 
energy of a diatomic molecule (E r ), we 
obtain the relation 

Er = -^~ (6) 

AT . 


where v is the frequency of rotation. 

On the other hand, if I denote the moment 
of inertia of the molecule rotating about its 
center of gravity, 

Er = \ {2-KVf 


Eliminating v from these two equations 
we should obtain an expression for the 
variation with the temperature of E r . The 
actual observations on the specific heat of 
hydrogen between the temperatures 35 and 
273 deg. K.§ were found to be quite different 
from those expected on the basis of this 

Einstein and Stern** have therefore sug- 
gested that instead of (6) we ought to write 
the equation for E r thus: 



E r=- h 

' kv 




assume the existence of an 

average latent energy — which is possessed 

by the rotating molecule even at the absolute 
zero. In this assumption they are in accord 

t According to Keesom (see reference p. 243) the specific heat 
of all gases at very low temperatures varies as the third power 
of the absolute temperature, so that this value 3 for the specific 
heat of hydrogen is only constant over a certain range of tem- 

§ A. Eucken. Berl. Akad. Ber. 1912. 141. 
** Ann. d. Physik, 40. 551 (1913k 



with Planck's most recent modification of the 
quantum theory. Furthermore, the photo- 
electric effect and emission of electrons by 
X-rays seemingly lead to the conclusion that 
the electrons in a metal possess a similar 

latent energy whose magnitude is — for each 

degree of freedom. 

The formula deduced for the specific heat 
of gases on this assumption accords well with 
the experimental data obtained by Eucken. 
It is, however, only fair to state, that almost 
as good an agreement has been obtained by 
Ehrenfest* without introducing the idea 
of a residual energy. 

W. H. Keesom has also tried to apply the 
quantum theory to the translational energy 
of a gas.f He uses arguments analogous to 
those of Debye, that is, from the elastic 
properties of the gas (these can most readily 
be calculated in this case from the velocity 
of sound in the gas) he calculates a semi- 
fictitious maximum frequency, v m , and then 
derives a formula for the specific heat of gases 
which is similar to that given above for solids. 
From the observed measurements of the 
pressure of helium at very low temperatures 
he is also inclined to favor the assumption 
that a zero point energy exists. In other 
words, while the rate of increase of energy 
per degree (the specific heat) tends to decrease 
to zero as the temperature is lowered, it 

absolute zero, a latent energy whose magni- 

tude is — for each degree of freedom. But 

it must be stated that more facts will have to 
be obtained before it will be possible to draw 
any definite conclusions in this direction. 

















Fig. 4. 

Specific Resistance of Gold, Mercury and 
Platinum at Low Temperatures 









•ft & 




? Ai 

f0 . 


IT Lli^ 






Cv =Atoa*'C f*/£/&*iT X SfSCtrtC M&AT at CoH»TAfirr 


C&i*STf<S#'* 0£ 















































Fig. 3. Atomic Heats of Diamond, Aluminum and Lead 

does not necessarily follow that the total 
energy itself becomes zero at the absolute 
zero. There are a number of reasons for 
believing rather that there exists, even at the 

*Verh7d. D phys. Ges. 15. 451 (1913). 

t Phvsik. Ztsch 14. 663 (1914). 

{ Abh. d. D. Bunsen Ges. (1914), p. 246. 

Electrical Resistance 

It has been known for a number of years 
that the resistance of metals decreases with 
the temperature. More recently a large 
number of investigations have 
been carried out in order to 
obtain accurate data on the 
variation of the electrical re- 
sistance of pure metals at 
extremely low temperatures, 
and the results have led to 
far-reaching speculations on 
the mechanism of the con- 
duction of heat and elec- 
tricity in metals. 

Fig. 4 gives the results 
obtained by Clay and OnnesJ 
on the resistance of mercury 
(Hg), gold (Au) and platinum 
(Pi) at temperatures ranging 
from 100 deg. K. down to the 
very lowest temperature at- 
tainable. Fig. 5 shows the 
same results plotted on a 
much larger scale. The ordinate gives the 
ratio between the specific resistance at T to 
that of deg. C. (273 deg. K.). 

"Superconducting" State 

The behavior of mercury is specially 
noteworthy. At 4.3 deg. K., the resistance 



is 0.0021 of its value at 273 deg. K.; but at 
4.21 deg. the resistance decreases very 
rapidly to a value which is less than one- 
millionth of that at deg. C. As Onnes 
expresses it. the mercury becomes "super- 
conducting." Very recently Onnes has 
extended these results and finds that at 
2.45 deg. K.. the resistance is less than 
2X10" 10 of that at deg. C, the potential 
difference at the extremities of the mercury 
column (contained in a tube 0.004 mm 2 
cross-section) being only 0.56 microvolt 
when the current density in the mercury was 
1024 amperes per square millimeter .* 

The same phenomenon has also been 
observed by Onnes in the case of tin and lead. 
The sudden disappearance of the resistance 
in the case of tin takes place at 3.806 
deg. K., the ratio of the resistance at 3.8 
deg. K. to that at 273 deg. K being less 
than 10" 7 . 

Lead becomes "superconducting" when 
immersed in liquid helium (4.3 deg. K.) ; the 
change, however, from the ordinary to 
the superconducting state occurs between 
14 and 4.3 deg. K. All these metals show 
the existence of threshold values for the 
current density, that is, with current densities 
above a certain definite value, the super- 
conducting state does not occur. Using a lead 
wire 0.025 mm. 2 cross-section, the threshold 
value at 4.25 deg. K. was observed to be 
between 420 and 940 amperes per square 

Onnes applied this result to the production 
of " resistanceless " coils having a great 
number of windings in a very small space. 
and therefore possessing high inductance. 
It is well known that if a coil through which 
a current is passing is suddenly short- 
circuited, the rate at which the current in the 
coil decreases to zero depends upon the 
ratio of the resistance to inductance. Using 
coils of lead wire immersed in liquid helium. 
Onnes found that even after several hours, 
the current through the short-circuited coils 
had not diminished noticeably. 

So far the existence of a superconducting 
state has been noticed only in the case of the 
three metals, mercury, tin, and lead. Each 
of these metals could be obtained in an 
exceptionally pure form. Onnes believes 
that all the other metals will show similar 
behavior at temperatures near the absolute 
zero when they can be obtained in a suffi- 
ciently pure state. 

» Science Abstracts, A. p. 114 (1914). 
t Science Abstracts, A. p. 385. (1914). 

The behavior of the samples marked Pt-B, 
Au (111) and Au (1") must be ascribed to the 
predominant influence exerted by slight 
traces of other metals at these very low 

Failure of the Older Theory of Conduction 

The curve for gold in Fig. 3 is typical 
of the behavior of all the metals in tempera- 
ture range 273 deg. K. to 20 deg. K. The 
resistance tends to approach the value zero 
asymptotically much in the same way as the 
specific heat. This fact has led to attempts 
to apply the quantum theory to explain 
the variation in electrical resistance at low 

According to the electron theory of con- 
duction the specific conductance, ■=-, is given 

by the relation} 

' Rt 



2 mu 



N = number of free electrons per unit 

L = mean free path 

e =unit electric charge 

w=mass of electron 

u = average velocity of the electrons. 

The assumptions used in deriving this 
equation are that electric conduction is due 
to a convection of charged particles (elec- 
trons) and that the collisions between atoms 
and electrons are non-elastic. But by making 
the further assumption that the free electrons 
in a metal possess the same average kinetic 
energy as the molecules in a gas at the same 
temperature, that is, that 

/2 mu^ = '-r feT§ 


it is possible to deduce a relation between the 
thermal and electrical conductivities which 
is known as the Wiedemann-Franz law. This 
law states that the ratio of the electrical 
to the thermal conductivity is a constant 
for all metals at the same temperature and 
varies directly as the absolute temperature. H 
The agreement between the empirical law 
obtained by Wiedemann and Franz and the 
relation deduced by means of considerations 

} E. P. Adams. Proc. Am. Inst. El. Eng., SS, 1159-1233 (1913). 
N. Campbell, Modern Electrical Theory (1913), also 
J. P. Minton, General Electric Review. March. 1915, 
p. 204. 

5 The constant — k is sometimes denoted by the symbol a. 

t J. P. Minton, General Electric Review, March. 1915. 
p. 207. 



based on the electron theory of conduction 
was taken to be a signal confirmation of the 
accuracy of these views. But it has been 
shown by Lorentz and others that the assump- 
tion that the average kinetic energy of the 
electrons increases as the absolute tempera- 
ture leads to conclusions which are at variance 
with the known distribution formulas for the 
energy radiated from a black body.* 

Furthermore, on the basis of the ordinary 
theory it is difficult to explain why the 
kinetic energy of the electrons should not 
exert an effect on the observed specific heats. 
Thus, if the number of free electrons is 
assumed to be the same as the number of 
atoms, and each electron is assumed to 

possess an average kinetic energy of — kT, 

the observed specific heat per gram-atom 

should be (6+tj) Nk = 7.5 calories, a con- 
clusion which is not at all in agreement with 
the observed values. (See section 2 above). 

Wien's Modification of the Electron Theory of 

These difficulties have led physicists in the 
past couple of years to discard the assumption 
expressed by equation (10). This naturally 
leads to the rejection of any theoretical basis 
for the Wiedemann-Franz law. Since equation 
(9) is merely an expression of Ohm's law in 
terms of the electron theory, it is taken as the 
starting point of the new theory which seeks 
to explain the observed variations in the 
specific resistance of metals as the temperature 
is lowered. 

While according to the older theory of 
Drude and Riecke, the value of the ratio 
N/u was assumed to vary with the tempera- 
ture, Wien assumes that this ratio remains 
independent of the temperature, while the 
mean free path, L, is the only quantity that 
does change. We know that in a metal the 
atoms vibrate about equilibrium positions 
which are arranged in regular lattice forms. 
The electrons travel between the rows of 
these atoms. The observed values of the 
specific heats show us that at very low tem- 
peratures the vibration of the atoms becomes 
extremely small in amplitude, and the 
number of vibrating atoms decreases rapidly. 
Thus, an electron starting out under the 
influence of even a small electric force, can 
travel a big distance without suffering 
collision with an atom; that is, the resistance 

* S. Dushman. General Electric Review, Sept., 1914. 

appears extremely small. At higher tem- 
peratures the atoms begin to vibrate more 
and more, so that the mean free path of the 
electron between collisions decreases; the 
kinetic energy of the electrons appears as 
Joule's losses in the conductor. 

As this theory requires that the kinetic 
energy of the electrons should be independent 
of the temperature, it is necessary to assume 
the existence of this energy even at T = 0. 
Here then we have again the conclusion which 
has been mentioned above, that there exists 
a zero point energy for the electrons in a 

Assuming that the mean free path, L, 
varies inversely as the square of the amplitude 
of the vibrating atoms, Wien deduces a 
relation of the form 





where A is a constant, and / is a definite 
function of the ratio between the quantity 
Q = pv m and T. It will be remembered that 
the quantity has already been referred to 
in section (2) as Debye's "Characteristic" 

At very low temperatures, the equation 
reduces to 

Rt = BT 2 (12) 

and at very high temperatures, it becomes 
of the form 

Rt = CT (13) 

A, B, and C are constants which vary for 
different metals. 

Over an intermediate range of temperatures 
the following relation holds fairly well: 

Rt _ T ( 1 | e \ ,,9 , . 
Rm ^273 "^ 298,000,/ /4 273 v ; 

This is in agreement with the observation 
that the specific resistance of most metals 
varies linearly with the absolute temperature. 
The values of the temperature coefficient 
of the resistance calculated from this equation 
are, however, found by Wien to be uniformly 
higher than the observed values (which range 
around 0.004 for most metals). 

Corresponding States 

The occurrence of the quantity Q = f$v m 
in the expression for the electrical resistance 
as a function of the temperature shows the 
existence of an intimate relation between the 
specific heat and electrical resistance. E. 
Gruneisenf has drawn attention to this 

t Verh. d. D. physik. Ges. 15, 186 (1913). 



relation and has, in fact, deduced the empirical 

R j=KC P (15) 

where C P denotes the specific heat per gram- 
atom at constant pressure, and K is a con- 




\ 0.0/0 








Vl/ III 

.11 „ 



f ; 











Fig. 5. Curves same as shown in Fig. 4, drawn on magnified 

scale to illustrate the "superconducting" 

state of mercury- 

It has already been observed that the 
function 9 has been designated by Debye 
as the "characteristic" temperature. The 
atomic heat of all monatomic substances is 
a unique function of the ratio &/T. In a 
similar manner we find that according to 
Wien Rt T 2 multiplied by a constant (which 
varies with each metal) is a unique function 
of 6,7'. 

The ratio 8 T thus plays the same role 
in connection with the properties of mon- 
atomic solids at low temperatures, as the ratio 
7 t T does in the consideration of the proper- 
ties of gases and liquids, where T c denotes 
the critical temperature. There is no doubt 
that in the near future it will be possi- 
ble to develop an equation of state for 
solids which will be quite as general as 
the conclusions which Van der Walls has 
developed for the transition from the gaseous 
to liquid state, and just as the ratio T c /T 
is of importance in considering corresponding 
states of gases and liquid, so the ratio 9 T 
must be the standard of reference for con- 
sidering the properties of solids at low 

Fig. 6* is of interest in this connection as it 
represents an attempt to correlate the 
electrical resistance with the values of 9 = j3i^ m 
derived from specific heat determinations. 

* H. Schimark. Ann. d. Physik, jfi, Tort I 1914). 

There appears to be a general tendency for 
the resistance to decrease with increasing 
value of jiv m . 

Difficulties in Wien's Theory- 
There are, however, some difficulties that 
prevent us from completely accepting Wien's 
theory as it stands. The Wiedemann-Franz 
law remains completely unexplained, and it is 
difficult to see how else it can be explained 
except by assuming that at least a fraction 
of the electrons possess an average kinetic 
energy which is the same as that of the 
molecules of a gas. Furthermore, as well 
known, the electron emission from hot 
metals increases exponentially with the tem- 
perature as shown by Richardson. This points 
to the conclusion that the kinetic energy 
of the electrons must also increase with the 
temperature — at least at the higher tem- 

On the other hand, Wien's theory does 
accord quantitatively with the experimental 
data at temperatures down to 20 deg. K. 
(No theory seems to have been advanced 
to account for the superconducting state.) 
Keesom has therefore attempted to reconcile 
the observations at low temperatures with 
those at higher temperatures by applying 
the conclusions which he has deduced regard- 
ing the specific heat of monatomic gases. 
He considers the electrons in the metal at 
very low temperatures as similar in all 
respects to a monatomic gas under the same 
conditions, and applies the quantum theory to 
calculate the variation in the concentration of 
free electrons with the temperature. He finds 
in this manner that at higher temperatures 
the average kinetic energy of the electrons 
is that demanded by the Law of Equi- 
partition and the electron emission must 
therefore vary according to a law which 
is approximately the same as that deduced 
by Richardson. With decreasing tempera- 
ture, the number of free electrons decreases 
and tends to a constant value at extremely 
low temperatures. In this region the velocity 
also tends to a constant value. The theory 
advanced by Keesom thus leads to conclu- 
sions which are in accord with Wien's theory 
for low temperatures, and with the older 
theory at higher temperatures. 

Thermo-Electromotive Force. Peltier-Effect 

The thermo-electromotive force is due to 
the potential difference developed when 
electrons are transferred from the hot junction 
of two different metals to the cold junction. 


If Na and Nb denote the concentrations in 
the two metals A and B, the difference of 
potential at the surface of contact according 
to the older theory is 

T . 2 k „ , Na 

Vab= -V T l0S 'N- B (1,i) 

Keesom shows that this formula holds only 
at the higher temperatures. At low tem- 
peratures the potential difference is 

l'.4B = 2.52X10- 22 r 3 

\Nb 2 Na 2 ) 


That is, the rate of change of thermo- 
electromotive force with temperature tends 
to vanish as the absolute temperature is 
approached. The equation is in agreement 
with the observations of Onnes and Hoist 
at the temperatures of boiling helium. 

The Peltier effect corresponds to the work 
required to transfer unit electric charge 
across the junction of two metals containing 
different concentrations of electrons. Accord- 
ing to Keesom the amount of heat developed 
at very low temperatures owing to this 
effect should vary as the fourth power of the 
absolute temperature. 

It is interesting to note that according to 
the above conclusions, the electrical resistance, 
the specific heat, the thermo-electromotive 
force, the Peltier effect, and probably most 
of the other properties obey such very simple 
laws at temperatures near the absolute zero. 
Not only do the actual values of each of these 
tend to disappear at extremely low tem- 
peratures,* but their differential coefficients 
with respect to the temperature tend towards 
the limit zero in the same manner. This is in 
accord with the predictions made by Nernst 
from a consideration of the rate of change with 
the temperature of the total energy of mon- 
atomic solids, and shows in another way 
that there exists an intimate relation between 
the specific heats and the other properties 
of solids at very low temperatures. 

Langevin's Theory of Paramagnetism 

A large number of investigations have been 
carried out during recent years in Onnes' 
laboratorv, on the magnetic properties of 
different " substances at low temperatures. 
While a more complete discussion of our 
present theories on the nature of magnetism 
must be reserved for a future occasion, a few 
remarks on Langevin's theory of paramag- 

*This statement is not true of all the properties if it be 
assumed (see pp. 242 and 2+8) that a zero-point energy exists. 

netism is necessary in order to understand 
why so much effort has been spent in investi- 
gating the effect of temperature on the 
magnetic susceptibility. 

Langevin assumes that the magnetism 
induced in paramagnetic substances is due 
to an orientation of the elementary magnets 
(which may consist of electronic orbits) under 

| 1 1 | ! | 


I 1 1 i 

1 ] 

1 ' 1 

1 1 


1 I 



! /" 




7 | 









Metal i $v 






F» 1 
























110 160 ISO 200 220 2tO 260 273,1 

Fig. 6. Specific Resistance of Various Metals from 
273 deg. K. to 20 deg. K. 

the influence of the magnetic field, which is 
opposed by the ordinary vibration due to 
heat energy. At any temperature there 
results therefore a static equilibrium between 
the force due to the outside magnetic field 
and that due to thermal agitation. Langevin 
was able to deduce from this the relation 
which had been previously derived by Curie 
from experimental data, and which states 
that the paramagnetic susceptibility (X) is 
inversely proportional to the absolute tem- 
perature, that is, 

x =w (18) 

where C is a constant. 

Deviations from Curie's Law at Low Temperature 

Onnes and his collaborators have, however, 
found that substances which follow this law 
at higher temperatures may begin to deviate 
from it at lower temperatures in such a 
direction that the susceptibility is lower 
than that deduced from the observations at 
higher temperatures. In explanation of this 
Oosterhuisj has suggested that most probably 

t Phys. Zeitschrift, ; .(, 862 (1913). 



it is no more justifiable to apply in this case 
the laws of statistical mechanics than in 
that of specific heats. According to the 
quantum theory, the average energy of an 
oscillator with two degrees of freedom is not 
kT, as demanded by the principle of equi- 
partition, but 

hv . hv 

U = 


t kT_ X 


[Compare equation (8) above.] 

Oosterhuis therefore writes equation (18) 
in the form 

A r 


where U is determined from equation (19) 
and v refers to the frequency of vibration 
of the molecular magnet. 

It is interesting to observe that according 
to this equation the curves giving the relation 
between the reciprocal of the magnetic sus- 
ceptibility ( - ) and T are of the same form 


as those indicated in Fig. 1 for the function F; 
but the actual values of -^ are proportional 

to the quantity I F +-~" )■ That is, the value 

of = does not decrease indefinitely (as de- 
manded by Curie's law) but tends to approach 
a practically constant value which is equal to 

~ 1 — 1, as the temperature approaches that 

of absolute zero. 

// v 
The introduction of the term -=- is justified 

on the following basis. If the thermal agita- 
tion of the molecular magnets ceases at the 
absolute zero, there is no force tending to 
oppose that of the magnetic field, all the 
elementary magnets must therefore orient 
themselves under the action of this field, 
that is, it must be possible to attain magnetic 
saturation. The fact that such saturation 
is not even approximately attained is taken 
to indicate that the molecular magnets possess 
a latent energv which is independent of the 

temperature and has the value — '■ 

Zero-Point Energy 

As pointed out by Oosterhuis* and as 
mentioned above, the existence of a zero- 

* Science Abstracts. A, 1914. p. 59. 

point energy is therefore made probable by 
four different lines of investigations. 

(1) The change in the specific heat of 
hydrogen at low temperatures has been 
explained by Einstein and Stern in a satis- 
factory manner by this assumption. 

(2) Keesom has shown that a similar 
assumption seems to be required for the 
translational energy of gas molecules in 
order to explain the deviations from the 
ordinary gas laws of a helium thermometer 
at low temperatures. 

(3) The assumption of a zero-point energy 
has been shown by Keesom to be of great 
importance in the theory of free electrons 
in metals. Moreover by making this assump- 
tion the views of Wien and the conclusions 
of the older theory are reconciled. 

(4) The deviations from Curie's law 
observed at low temperatures may be corre- 
lated and quantitatively explained by this 

While it may be possible to explain each 
of these sets of observations without the 
necessity of introducing a zero-point energy, 
yet the cumulative effect of these four differ- 
ent lines of investigations is strongly in favor 
of the assumption that the molecules of a gas 
and the free electrons in a metal possess a 
latent energy which is independent of the 
temperature and persists even at the absolute 


The above are. briefly, some of the specu- 
lations and conclusions which have been 
suggested by the results of the investigations 
at low temperatures. The absolute zero 
appears to us from the point of view of the 
present theories as a sort of unattainable 
limit. As we approach this temperature 
more and more closely, the practical diffi- 
culties in the way "of obtaining still lower 
temperatures become greater and greater 
until finally they become insurmountable. 
All the criteria by which we ordinarily 
determine temperature gradually disappear 
and the absolute zero itself becomes, as it 
were, a will-o'-the-wisp which continually 
draws us on and yet remains just as remote 
from our reach. 

However, the investigations at low tem- 
peratures lead to this important conclusion : 
that the properties of all substances tend to 
obey very simple laws under these conditions, 
and furthermore, that all these properties 
are probably connected by functions which 
are perfectly general and valid for all sub- 



By A. A. Ross 

Railway and Traction Engineering Department, General Electric Company 

The following article contains information which should prove to be of much practical value to those 
interested in the operation of railway motor gears and pinions. After locating and identifying those causes 
which have produced unsatisfactory operation, the author discusses each one and formulates such recommenda- 
tions as will lessen the premature breakage and excessive wear of bearings and teeth. The limitations of 
motor design as influenced by the use of standard 14^2 deg. pressure angle and 20 deg. stub pressure angle 
teeth are denned. The most beneficial relative hardness of gear to pinion is named, and a description of good 
and bad methods of mounting and dismounting pinions is given. The article is concluded by a discussion 
and statement of reasonable gear mounting pressures. — Editor. 

The limitations imposed upon the space 
available for gears and pinions in the design 
of modern railway motors, and the severe 
conditions under which railway motor gears 
and pinions are operated, necessitate the 
use of materials which will insure protection 
against breakage and secure the maximum 
resistance to abrasive wear. While the gear 
and pinion manufacturers have been striving 
to meet these conditions with various grades 
of steel and special methods of treating the 
steel, very few operators have taken steps 
to improve operating conditions. 

Apparently the average operator has never 
given this side of the question serious con- 
sideration, a pinion is a pinion and a gear is 
a gear, and if they break or wear out rapidly 
it is simply defective material and up to the 
manufacturer to make good. If the manu- 
facturer does not feel so inclined the operator 
invariably changes to some other manu- 
facturer's product, and in nine cases out of 
ten the original trouble recurs. Had the 
operating conditions been definitely known, 
the trouble might have been overcome with 
the original material and the operator, the 
manufacturer and the trade benefited there- 

b y- 

The mere replacement of a gear or pinion 
covers a very small percentage of the actual 
cost to the operator. The writer will endeavor 
to briefly outline these conditions in the 
following order: 

Motor design limitations. 
Variation in gear and pinion life. 
Operators' responsibility for broken gears 
and pinions. 

Motor Design Limitations 

The involute or single curve tooth is best 
suited and most commonly used for railway 
motor work, for two reasons; first, on account 
of the greater thickness at the root of the 
tooth, and second, because the distance 

between the centers of the gear and pinion 
can be slightly increased without seriously 
affecting the mesh of the teeth. 

In city service 3 diametral pitch teeth are 
the most popular, with an occasional appli- 
cation of 3 }/2 and 4 diametral pitch on small 
motors for 24-inch wheel equipments. For 
heavy high speed duty, such as interurban, 
suburban and locomotive service, the size 
of teeth is usually either 2}^, 2)4,, 2 or \% 
diametral pitch. For the smoothest operation 
the teeth should conform to the standard 
14}/2-degree pressure angle, but on account 
of high tooth loads the motor designer is 
frequently forced to use the 20 degree pressure 
angle stub tooth. 

A comparison of the 2^ pitch standard and 
20 degree angle stub teeth is shown in Figs. 
1, 2 and 3. The contours of the teeth in the 
layouts are theoretically correct and may 
vary slightly from the actual teeth, but the 
working contacts are sufficiently correct for 
comparisons. They certainly will not improve 
as the teeth become worn. It will be noted 
that the diameter of the base circle from which 
the involute or face radius of the tooth is 
generated decreases as the pressure angle 
increases. Consequently, the tooth with the 
20 degree pressure angle is much thicker 
at the base. The "Whole Depth" of a stub 
tooth is usually made equal to the next half 
size smaller standard tooth; that is, the 
addendum or the distance from the pitch 
diameter to the top, and root or the distance 
from the pitch diameter to the base of 2}/% 
pitch stub corresponds to the 3 pitch stand- 
ard; the 2 pitch stub to the 2Yi pitch 
standard, etc., but the pitch diameter, the 
tooth thickness at pitch line, the face and 
flank radii of the tooth are the same for both 
standard and stub. 

The stub tooth is from 40 to 50 per cent 
stronger than the standard, but the mesh or 
working contact of the teeth is not so desir- 





Fig. 1. Four Different Mesh Positions of a 2H Pitch, 18 Tooth Pinion, and 68 Tooth Gear 



Fig. 2. Angle of Action of the Teeth shown in Fig. I 


Fig. 3. Angle of Action of the Teeth shown in Fig. 1, but with Distance between Gear Centers Increased 



Fig. 1 shows a 23-2 pitch, 18 tooth pinion 
and a 6S tooth gear standard and stub in 
mesh in four different positions. 

Fig. 2 shows the angle of action of this 
combination or the angle through which the 
pinion passes while a tooth is in contact. 
Teeth Nos. 1 and 1' are two successive teeth. 
No. 1 is just coming in contact and 1 and 1' 
travel to the 2 and 2' positions before 1' or 2' 
as shown disengages. Therefore between 
1 and 2 positions there are two teeth in 
contact and between 2 and 1' positions only 
one tooth is in contact. It is obvious at a 
glance that the working contact of the 
standard is much better than the stub and 
that the contact of the stub tooth in thel' 
position is at the extreme top or excessive 
friction point before it is relieved by the 
incoming No. 1 tooth. 

In Fig. 3 the distance between the centers 
of the gear and pinion have been increased 
to represent the conditions when the armature 
linings are worn yj in. and the axle lining 
about yg in. No. 1' tooth is about to disengage 
as No. 1 engages. While the working contact 
is bad on the standard, it is considerably 
worse on the stub. 

Comparative tests which the writer has 
followed, and reports from operators, would 
indicate that 20 deg. stub tooth gearing is 
more noisy than the standard. No doubt this 
is due to the inferior mesh. 

The increased angle of pressure will 
increase the pressure on the bearing linings 
but the increase is so slight that the effect 
on the life of the linings is negligible. 

The reader will, no doubt, ask the ques- 
tions: Why adopt the 20 deg. angle stub 
tooth since it does not give the best servicer 
Why not gain the required strength by 
choosing a larger standard tooth ? In this the 
designer is usually limited to a ratio suitable 
to meet service requirements. To maintain 
this ratio and adopt a larger tooth is very 
often impossible for the following reasons : 
First. It would mean an increase in the 
distance between the center of the gear and 
pinion which adds to the weight and price 
of the motor, both of which would be seri- 
ously objectionable to the operator. 
Second. The design of the trucks may not 
permit an increase in the distance between 
the center of the axle and suspension side 
of the motor. 
Third. The minimum number of teeth in the 
pinion is limited to the diameter of the 
shaft and thickness of the section of metal 
between the base of the tooth and bottom 

of the keyway. Consequently the larger 
tooth invariably means a considerable in- 
crease in the outside diameter of the gear. 
Such an increase may not be possible on 
account of the wheel diameter and track 
clearance limits. 

Regardless of the mesh there are large 
quantities of stub tooth gearing being oper- 
ated throughout the country, and with the 
exception of the complaint on noise, they are 
giving perfect satisfaction. Apparently, the 
difference in total life has not been noticed. 
However, the writer would recommend the 
use of the standard tooth wherever it is 
possible to apply it, for railway motor 
gearing is bad enough at the best. If the use 
of stub teeth is imperative, limit the practice 
to teeth whose "whole depth" is not less 
than the 3 pitch standard, for the average 
operator will allow the same limits for lining 
wear regardless of the length of the teeth 
in his gearing. If the length of the teeth in 
Fig. 3 were reduced the working contacts 
would be still worse. 

Variation in Gear and Pinion Life 

From actual service observations it is 
evident that in straight carbon steel or 
non-alloy steel the harder the wearing 
surface of the tooth, the greater the resistance 
to abrasive wear. Case hardened material 
now offered to the trade under various trade 


Fig. 4. Sand shown in Vial Represents One Per Cent of Sand 
by Weight in 12 Lb. of "Gear Lubricant 

names by practically all gear and pinion 
manufacturers, affords about the hardest 
possible tooth surface, but it does not afford 
maximum protection against breakage. While 
not wishing to offer excuses for manufacturing 


defects, the structure of this steel with its 
glass hard brittle surface is very susceptible 
to injury from shocks such as may be trans- 
mitted to the teeth during motor flashovers 
or when at high speed the wheels hit high 
rail joints, frogs, etc. ; but the greatest source 
of danger lies in the operator's methods of 

mum protection against breakage and a 
high uniform hardness which resists abrasive 
wear almost to the same degree as case 
hardened material, it devolves upon the 
operator to determine by actual service 
tests on his own equipment which is the most 






' 4 

. ■ <£' '.* 



• f* 




^ i • 



* '* 



?t^* *» 





' '# ■.-■•"■ 







Grease Containing 2 Per Cent of Sand by Weight 

Grease Containing 5 Per Cent of Sand by Weight 




Grease Containing 10 Per Cent of Sand by Weight 

Fig. 5. Photo-micrographs (26 dia.) of a Popular 

mounting and dismounting; this will be 
referred to later. The material is also very 
expensive to manufacture, but its total life 
has been so much greater than the old 
combination which consisted of oil treated 
pinions and untreated cast steel gears that 
its first cost and breakage has been overlooked 
by the operator. 

"With the advent of the less expensive 
specially treated homogeneous steel having 
physical characteristics which afford maxi- 

Grease Containing 25 Per Cent of Sand by Weight 
Motor Grease Containing Various Percentages of Sand 

It is unsafe for one operator to use the life 
values established on some other road for the 
life is in a great measure affected by the 
ratio. Still it is quite common to find a 
radical variation in the life of the same 
grades of gears and pinions on two roads 
which have duplicate equipments. Such 
variations can only be accounted for as 
follows : 

The first and greatest factor is the grit or 
cutting substance which accumulates in the 



■xy ajar 


T ^J i JH M S3*A*s4*S SSNfKf? 




|M- L .-#-t 'S § 


gear pan and when mixed with the lubricant 
acts as an abrasive lap on the gear, pinion 
or both. Practically every Master Mechanic 
will disclaim its presence, but it is there and 
it has been found in quantities as high as 
24 per cent. 

It is doubtful if the reader appreciates what 
even one per cent of sand really means. 
The average amount of lubricant in gear pans 
is about 12 pounds. The quantity of sand 
in the vial, Fig. 4, represents 1 per cent of 
sand in 12 lb. of lubricant. The vial is 1% 
inches in diameter. Pause for an instant 
and consider what 10 per cent or 10 times, 
24 per cent or 24 times, this quantity means. 

Fig. 5 shows photo-micrographs of a 
popular motor grease containing various 
percentages of sand magnified to 26 diam- 
eters. The sand used in the mixture was first 
screened through an SO-mesh screen. 

The sand usually enters between the gear 
hub and gear pan in the form of street dust, 
brake shoe dust and wheel wash. Very often 
the contour of the web and hub of the wheel 
is such that the natural flow of the wheel 
wash is against the opening in the gear pan 
as shown in Fig. 7. It would be much better 
if this were retarded by either making the 
diameter of the wheel hub larger or smaller 
than the gear hub. The clearance between 
the gear hub and gear pan is sometimes 
enclosed by a felt dust guard, but this is not 
a permanent protection as the felt soon 
fills up with sand, and breaks off. Careless- 
ness when adding lubricant or when the 
lower half of the pan is lying in the pit 
during inspection is another source of dirt 
getting into the grease. One of the largest 
operators in the east traced the cause of rapid 
wear to the presence of sand in the lubricant 
which was carried into the pan by the wheel 
wash during a period of heavy snow and slush. 
The sand or stone dust seems to scour 
the lubricant off the teeth. 

Before shipment the finished surfaces of 
gears and pinions are given a coating of 
slushing compound to prevent rust. This 
should be removed before the gears and 
pinions are placed in service as it becomes 
filled with lime and sand during shipment. 

The second factor is excessive lining wear 
which permits improper mesh as shown 
in Fig. 3. The design of the motor and the 
truck prevents the use of a bearing on each 
side of the gear and pinion so that when 
power is transmitted from the pinion to the 
gear both the armature and axle shafts tend 
to spring diagonally away from one another. 

The armature shaft bearing adjacent to the 
pinion becomes worn on the side which is 
farthest from the axle, while the bearing 
on the other end wears on the side near the 
axle. The wear on both bearings is not 
radial with the center of the axle but at t w. i 

Fig. 7. Diagram showing the Natural Direction of Flow of 
Wheel Wash Against the Gear Cover 

points about 45 degrees above and below the 
radial line, according to the direction of 
rotation. This allows the pinion teeth to set 
at an angle to the gear teeth, which means 
that the ends of the teeth next the motor 
receive the greatest percentage of shocks and 
must perform the major portion of the work. 
Such conditions account for the greater wear 
on the motor ends of the teeth. Consequently, 
very little would be gained by increasing the 
present standard width of face. Furthermore, 
the axle linings wear on the upper portion 
of the bore away from the armature shaft 
which tends to carry the pitch lines of the 
gear and pinion still further apart and forces 
the working contacts to take place on the 
top or excessive friction points. 

Some operators claim that it is impossible 
to maintain J^j in. as the limit for axle lining 
wear, as it forces too frequent renewals 
and that the slight reduction in gearing 
maintenance is offset by an increase in lining 
maintenance. If gearing maintenance were 
the only consideration there might be some 
ground for the argument, but the improper 
mesh also produces a noisy chattering, which 
effects commutation and the nerves of the 
population living along the route over which 
the gearing is operated. If the operator doubts 
the effect on the rest of the equipment -let 
him get into the pit under a motor with the 
axle linings worn say ^ in., start the car so 
that the pinion climbs the gear and, as the